Hu_Moments_.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [],
      "mount_file_id": "1NM3I-Tl5UyDlWTfGQgsB-ekgsecx_o7x",
      "authorship_tag": "ABX9TyPK83OYyjUhzXnAYoZrdWjp",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/RuthKassahun/Skin-Lesion-Classification/blob/main/Hu_Moments_.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# HU moments with Hair Removed "
      ],
      "metadata": {
        "id": "uZZBnPHNLLVZ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Import Libraries"
      ],
      "metadata": {
        "id": "VEyKRvtkLVv6"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 182,
      "metadata": {
        "id": "C6XWF4DqCFRw"
      },
      "outputs": [],
      "source": [
        "import numpy as np \n",
        "import matplotlib.pyplot as plt\n",
        "import glob\n",
        "import cv2\n",
        "import os\n",
        "import seaborn as sns\n",
        "import pandas as pd\n",
        "from skimage.filters import sobel\n",
        "from skimage.feature import greycomatrix, greycoprops\n",
        "from skimage.measure import shannon_entropy\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.svm import SVC, LinearSVC\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.ensemble import AdaBoostClassifier\n",
        "from sklearn.preprocessing import LabelEncoder, StandardScaler, MinMaxScaler\n",
        "from sklearn import metrics\n",
        "from pathlib import Path\n",
        "import xgboost as xgb\n",
        "import lightgbm as lgb\n",
        "from PIL import Image, ImageOps\n",
        "from sklearn.model_selection import GridSearchCV, StratifiedKFold\n",
        "from sklearn.metrics import accuracy_score, auc, roc_curve, roc_auc_score, recall_score, confusion_matrix\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Load Data"
      ],
      "metadata": {
        "id": "6fxJw8o3LfvZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(os.listdir(\"/content/drive/MyDrive/train/sample_train\"))\n",
        "SIZE = 224"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XxVgvMwHCrME",
        "outputId": "a7313be2-a68a-431c-f8d5-80368a79d1ac"
      },
      "execution_count": 183,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['other', 'nevus']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from zipfile import ZipFile\n",
        "file_name = '/content/sample_check.zip'\n",
        "\n",
        "with ZipFile(file_name, 'r') as zip:\n",
        "  zip.extractall()\n",
        "  print('Done')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gG-aShBOxbjk",
        "outputId": "2959ebdc-c0b4-4d5d-e5df-16e889d04f30"
      },
      "execution_count": 184,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Done\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def extract_hu_moments(img):\n",
        "   \n",
        "    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
        "    feature = cv2.HuMoments(cv2.moments(gray)).flatten()\n",
        "    return feature"
      ],
      "metadata": {
        "id": "oGuoxk8GIz_K"
      },
      "execution_count": 185,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_images = []\n",
        "train_labels = [] \n",
        "hu_features = []\n",
        "#for directory_path in glob.glob(\"cell_images/train/*\"):\n",
        "for directory_path in glob.glob(\"/content/drive/MyDrive/CAD_PROJECT/hair_removed/train/*\"):\n",
        "    label = directory_path.split(\"/\")[-1]\n",
        "    #print(label)\n",
        "    for img_path in glob.glob(os.path.join(directory_path, \"*.jpg\")):\n",
        "        #print(img_path)\n",
        "        img = cv2.imread(img_path)[:, :, :3] #Reading color images\n",
        "        img = cv2.resize(img, (SIZE, SIZE)) #Resize images\n",
        "        train_images.append(img)\n",
        "        train_labels.append(label)\n",
        "        hu_feature = extract_hu_moments(img)\n",
        "        hu_features.append(hu_feature)\n",
        "        \n",
        "train_images = np.array(train_images)\n",
        "train_labels = np.array(train_labels)"
      ],
      "metadata": {
        "id": "LMC30n42p6zS"
      },
      "execution_count": 224,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_images"
      ],
      "metadata": {
        "id": "-qbE1nYvycKH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_labels)"
      ],
      "metadata": {
        "id": "u-QOj9ELwCdk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3368a45c-db5b-400c-dddb-d9b101bf3ee0"
      },
      "execution_count": 226,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['nevus' 'nevus' 'nevus' ... 'other' 'other' 'other']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_images = []\n",
        "test_labels = []\n",
        "hu_features_test = []\n",
        "\n",
        "#for directory_path in glob.glob(\"cell_images/test/*\"): \n",
        "for directory_path in glob.glob(\"/content/drive/MyDrive/CAD_PROJECT/hair_removed/test/*\"):\n",
        "    test_label = directory_path.split(\"/\")[-1]\n",
        "    for img_path in glob.glob(os.path.join(directory_path, \"*.jpg\")):\n",
        "        img = cv2.imread(img_path)[:, :, :3]\n",
        "        img = cv2.resize(img, (SIZE, SIZE))\n",
        "        test_images.append(img)\n",
        "        test_labels.append(test_label)\n",
        "        hu_feature_test = extract_hu_moments(img)\n",
        "        hu_features_test.append(hu_feature_test)\n",
        "        \n",
        "        \n",
        "test_images = np.array(test_images)\n",
        "test_labels = np.array(test_labels)\n"
      ],
      "metadata": {
        "id": "OmGYCrkIqj9i"
      },
      "execution_count": 227,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Label Encoding"
      ],
      "metadata": {
        "id": "CerVmwjfLvEs"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import preprocessing\n",
        "le = preprocessing.LabelEncoder()\n",
        "le.fit(test_labels)\n",
        "test_labels_encoded = le.transform(test_labels)\n",
        "le.fit(train_labels)\n",
        "train_labels_encoded = le.transform(train_labels)"
      ],
      "metadata": {
        "id": "Nx-a7VFwrE3R"
      },
      "execution_count": 228,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "x_train, y_train, x_test, y_test = hu_features, train_labels_encoded, hu_features_test, test_labels_encoded\n"
      ],
      "metadata": {
        "id": "pPcS7NIareJw"
      },
      "execution_count": 229,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_for_ML = x_train\n",
        "X_test_ML = x_test\n"
      ],
      "metadata": {
        "id": "xdDRRdCa9q2r"
      },
      "execution_count": 230,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## SVM"
      ],
      "metadata": {
        "id": "mRt5SBmbMCBH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pipe1 = Pipeline([\n",
        "                 (\"scale\", StandardScaler()),\n",
        "                 (\"classifier\", SVC(class_weight='balanced', probability=True, kernel= 'rbf', gamma=0.01, C=100)) #radial basis function kernel\n",
        "                 ])"
      ],
      "metadata": {
        "id": "8NeblTA2AdM3"
      },
      "execution_count": 231,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pipe1.fit(X_for_ML,train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jt-exYkgAeb4",
        "outputId": "062a58a4-9a79-4469-c4c6-8f975de12e87"
      },
      "execution_count": 232,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Pipeline(steps=[('scale', StandardScaler()),\n",
              "                ('classifier',\n",
              "                 SVC(C=100, class_weight='balanced', gamma=0.01,\n",
              "                     probability=True))])"
            ]
          },
          "metadata": {},
          "execution_count": 232
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Test set evaluation\n",
        "prediction = pipe1.predict(x_test)\n",
        "# Visualization\n",
        "cmat = metrics.confusion_matrix(prediction,test_labels_encoded)\n",
        "s = sns.heatmap(cmat,annot=True,xticklabels=label, yticklabels=label)\n",
        "s.set(xlabel='Predicted', ylabel='Actual')\n",
        "acc = cmat.trace() / cmat.sum()\n",
        "print('Accuracy: {0:5.2f}%'.format(acc*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "_Ycj94-_CCiT",
        "outputId": "2da9102c-3155-49a8-c697-6f5f3a1835a5"
      },
      "execution_count": 233,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: 49.48%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEKCAYAAAAYd05sAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAfS0lEQVR4nO3deZhdVZnv8e+PhFGpJBDSIiUGGhBBuqMEiKbhRmgQk5gEoSFcEJAhDFeJ0BJIXy48IN4r4oCogGWUocPUEknoMAQhRhMkgUwyBWWShgA3IoEgAknqvP3HXgWHyqk6u1J16lTt+n141pPaa0/vjj5vray19tqKCMzMrHg2qXcAZmZWG07wZmYF5QRvZlZQTvBmZgXlBG9mVlBO8GZmBeUEb2ZWA5L6SVomaXbaPlDSUkmPSrpOUv82zjte0pOpHF9Wv7ekRyQ9JekKSaoWgxO8mVltTAZWAEjaBLgOmBgRnwCeA45vfYKkbYALgf2AfYELJQ1Ku68CTgF2TeXQagE4wZuZdTFJjcAYYFqq2hZYGxF/TNu/Ag6vcOrngF9FxKsRsTodd6ik7YGGiFgY2dup1wMTqsVR8Z8IPcG6V57xK7YFtungnesdghVL1e6KajqSczbb7u9PBSaVVTVFRFPZ9uXAFGDrtP0K0F/S8IhYDBwBfKTCpXcAni/bfiHV7ZB+bl3frh6b4M3MeqqUzJsq7ZM0FlgVEUskjUrHh6SJwPclbQ7cAzTXOk4neDMzgFKX5duRwDhJo4EtgAZJ0yPiWGB/AEmHALtVOHclMKpsuxGYl+obW9WvrBaI++DNzACa1+cv7YiIqRHRGBFDgYnA3Ig4VtIQgNSCPxe4usLpc4BDJA1Kg6uHAHMi4iVgjaQRafbMccCsao/kBG9mBkSUcpeNdI6kFcDDwH9GxFwAScMlTctiiFeBbwAPpXJxqgM4g2zQ9ingaeCuajdUT10u2IOsxeZBVutinR5kXfvCI/kHWRv36vT9uoP74M3MADa+Zd5jOcGbmUFXDrL2GE7wZmbgFryZWVFFldkxvZETvJkZQMkteDOzYnIXjZlZQXmQ1cysoNyCNzMrKA+ympkVlAdZzcyKKcJ98GZmxeQ+eDOzgnIXjZlZQbkFXzzNzc0cddKZDNluMFdedhGLliznOz+axrp169njY7tw8dSz6N+/X73D7DJ97XnNcmteV+8Iulyf/+DH9F/MYuehOwJQKpX4t0u+y2UXncfM6Vfz4Q8NYdZd99Y5wq7V157XLLdSKX/pJfp0gn951Z/57e8e5PAvfA6A115fw6b9+zN0x+zTh5/e51PcO29BPUPsUn3tec06JEr5Sy9R8wQv6R8lfSWVf6z1/Tri0h/8hLPPOAkp+2sYNHAAzc0lHl3xRwDumbeAl1e9Us8Qu1Rfe16zDnELvmMkTQZuAIakMl3SV2t5z7zm3b+IbQYNZM/dd323ThKXXXwe376iiYknT+YDW23JJpsU4x85fe15zTqsixO8pH6SlkmanbYPkrRU0nJJCyTtUuGcY9L+llKSNCztmyfpD2X7hlSLodaDrCcB+0XEmynAS4EHgB9WOljSJGASwJXfvYSTjzu6ZoEte/hx5i1YyPwHHuKdtet4882/ce5F3+bSC6dw/VXfAeD+RUt47vmVNYuhO/W15zXrqOj6QdbJwAqgIW1fBYyPiBWSzgDOB054XwwRN5A1ipG0FzAzIpaXHXJMRCzOG0CtE7yA8tfDmmnn47gR0QQ0Qe0/un3W6V/mrNO/DMCDSx/m2ptmcOmFU/jL6tfYdtBA1q5dy89v+AWTjp9YyzC6TV97XrMO68K+dUmNwBjgm8DZLXfgvWQ/AHixymWOBm7uTBy1TvDXAIsk3Za2JwA/q/E9O+WaG27lN797kCiVOOqwMey397B6h1RTfe15zdrUtX3rlwNTgK3L6k4G7pT0FrAGGFHlGkcB41vVXSOpGZgBXBIR7TaEVWV/p0n6FPBPaXN+RCzLc16tW/BWX5sO3rneIVixtNkzkNdb916dO+dsdfDpp5K6k5Om1AOBpLHA6Ig4Q9Io4OsRMVbSL4FLI2KRpHOAj0XEyZWuL2k/YFpE7FVWt0NErJS0NVmCnx4R17cXZ81fdIqIpcDSWt/HzKxTOtCCL+9OrmAkME7SaGALoEHSHcDuEbEoHXMLcHc7t5gI3NTqnivTn29IuhHYF2g3wXvKhJkZdNk8+IiYGhGNETGULFHPJetqGSBpt3TYwWQDsBtQNo/5SMr63yX1lzQ4/bwpMBZ4tNoj9fmlCszMAFhfuw9+RMR6SacAMySVgNXAiQCSxgHDI+KCdPgBwPMR8UzZJTYH5qTk3g+4F/hptfvWvA9+Y7kPvtjcB29drPN98LO/lzvnbDn27E7frzu4BW9mBr3qDdW8nODNzKBXrTGTlxO8mRm4BW9mVlhuwZuZFVQNZ9HUixO8mRlAD51R2BlO8GZm4D54M7PCcoI3MysoD7KamRVUc3P1Y3oZJ3gzM3AXjZlZYTnBm5kVlPvgzcyKKUqeB29mVkzuorFa2PLD+9c7hG63fu3Keodg9n6eRWNmVlBuwZuZFZQTvJlZQRVwsbFN6h2AmVmPUCrlLzlI6idpmaTZafsgSUslLZe0QNIuFc4ZKumtdMxySVeX7dtb0iOSnpJ0haSq34V1gjczAyhF/pLPZGBF2fZVwDERMQy4ETi/jfOejohhqZzW6vxTgF1TObRaAE7wZmaQzaLJW6qQ1AiMAaaVVQfQkH4eALyYNzRJ2wMNEbEwIgK4HphQ7Tz3wZuZAdGBQVZJk4BJZVVNEdFUtn05MAXYuqzuZOBOSW8Ba4ARbVx+J0nL0jHnR8R8YAfghbJjXkh17XKCNzODjnS9kJJ5U6V9ksYCqyJiiaRRZbvOAkZHxCJJ5wDfI0v65V4CdoyIv0jaG5gpac8OPMX7OMGbmUFXrkUzEhgnaTSwBdAg6Q5g94hYlI65Bbh7gxAi3gHeST8vkfQ0sBuwEmgsO7Qx1bXLffBmZtBlg6wRMTUiGiNiKDARmAuMBwZI2i0ddjDvH4AFQNJ2kvqln3cmG0x9JiJeAtZIGpFmzxwHzKr2SG7Bm5kBrK/dUgURsV7SKcAMSSVgNXAigKRxwPCIuAA4ALhY0jqgBJwWEa+my5wBXAtsCdyVSrsUPXRy/7pXnumZgdWA16Ix67Sqc8KrefP/HJk753zgG//R6ft1B7fgzcygQ4OsvYUTvJkZHZsm2Vs4wZuZgVvwZmaF5QRvZlZQ/uCHmVkx+ZusZmZF5QRvZlZQnkVTPM3NzRx10pkM2W4wV152EYuWLOc7P5rGunXr2eNju3Dx1LPo379fvcPsMptssgmLFt7FiytfZvxhxzNv7i/54NYfBGDIdtvy0OLlHH7ESXWO0qwOCtiC7/Nr0Uz/xSx2HrojAKVSiX+75LtcdtF5zJx+NR/+0BBm3XVvnSPsWmd+9WSeeOLJd7dHHfhFhu9zCMP3OYSFi5Zw28yqbz+bFVPXf/Cj7vp0gn951Z/57e8e5PAvfA6A115fw6b9+zN0x2zRtk/v8ynunbegniF2qR122J7Rnz+In//8pg32bb31B/nsqJHMmrXBAndmfUI0l3KX3qJPJ/hLf/ATzj7jJKTsr2HQwAE0N5d4dMUfAbhn3gJeXvVKPUPsUt/77kWcN/USShX6GsePP5S5v76fN974ax0iM+sB3ILPT9Kleepa7Z8kabGkxdOu37CV2ZXm3b+IbQYNZM/ddy2/P5ddfB7fvqKJiSdP5gNbbckmmxTjd+CY0f/MqlWvsHTZIxX3TzxyPDffMrObozLrOaIUuUtvUctB1oOBc1vVfb5C3bvKv5JS69Uklz38OPMWLGT+Aw/xztp1vPnm3zj3om9z6YVTuP6q7wBw/6IlPPd8MVY9/MxnhvOFsYfw+UMPZIstNqehYWuuu/YKjj/hTLbddhD77PNJDv+X1h+XMetDelHizqvLlwuWdDrZusU7A0+X7doauD8ijs1zne5cLvjBpQ9z7U0zuPKyi/jL6tfYdtBA1q5dy+lfv4BJx09kv72H1fT+3b1c8P844NOcfdZpjD/seAAmnfIlRozYmxNP+lq3xeDlgq2LdXr53te/dFDunDPg3+/rs8sF30i2EP3/A84rq3+jbOH6HuuaG27lN797kCiVOOqwMTVP7j3BUUeO49uX/bjeYZjVVazvPYOnefmDHz2AP/hh1mmdblG/dtRnc+ecgbf8us+24M3Mep3eNHiaVzGmiJiZdVapAyUHSf0kLZM0O20fJGmppOWSFkjapcI5B0taIumR9OeBZfvmSfpDOn+5pCHVYnAL3syMmrTgJwMrgIa0fRUwPiJWSDoDOB84odU5rwBfiIgXJX0CmAPsULb/mIhYnDcAt+DNzKBLW/CSGoExwLSy6uC9ZD8AeLH1eRGxLCJa6h8DtpS0eccfJuMWvJkZEOvzHytpEjCprKopvcfT4nJgCtn08BYnA3dKegtYA4yocpvDgaUR8U5Z3TWSmoEZwCVRZZaMW/BmZkCUOlAimiJieFl5N7lLGgusioglrW5xFjA6IhqBa4DvtRWLpD2BS4FTy6qPiYi9gP1T+VK1Z3IL3swMcg+e5jASGCdpNLAF0CDpDmD3iFiUjrkFqLiyX+reuQ04LiLefVk0IlamP9+QdCOwL3B9e4G4BW9mRsda8O1eJ2JqRDRGxFBgIjAXGA8MkLRbOuxgsgHY95E0ELgDOC8i7i+r7y9pcPp5U2As8Gi1Z3IL3syM6om7U9eOWC/pFGCGpBKwGjgRQNI4YHhEXAB8BdgFuEDSBen0Q4A3gTkpufcD7gV+Wu2+fpO1B/CbrGad1uk3S///qFG5c87fzZvnN1nNzHqLWrbg68UJ3swMiFKvaJR3iBO8mRluwZuZFVaEW/BmZoXkFryZWUGVmt2CNzMrJA+ympkVlBO8mVlB9dB3PjulzQQv6Ydk6xdXFBFn1iSiZNPBO9fy8j2K3+o0q7++1oLP/dUQM7Perk9Nk4yI67ozEDOzemrui7NoJG0HnAvsQba2MQARcWCbJ5mZ9TJFbMHnWQ/+BrJ1i3cCLgL+BDxUw5jMzLpdlJS79BZ5Evy2EfEzYF1E/CYiTgTcejezQonIX3qLPNMk16U/X5I0huxL4NvULiQzs+7Xm1rmeeVJ8JdIGgD8K/BDoIHs47FmZoXRXCreF0yrJviImJ1+fB34bG3DMTOrj97U9ZJXnlk011DhhafUF29mVgilLp5FI6kf2ftEKyNirKSDgMvIxj7/CpwQEU9VOG8qcBLQDJwZEXNS/aHAD8i+yTotIr5VLYY8XTSzy37eAjiMrB/ezKwwajBNcjLZDMSGtH0VMD4iVkg6AzgfOKH8BEl7ABOBPYEPA/dK2i3t/jFwMPAC8JCk2yPi8fYCyNNFM6NVADcBC6qdZ2bWm3RlF42kRmAM8E3g7JZb8F6yH0DlhvJ44OaIeAd4VtJTwL5p31MR8Uy6/s3p2M4l+Ap2BYZsxHlmZj1WR7poJE0CJpVVNUVEU9n25cAUYOuyupOBOyW9BawBRlS49A7AwrLtF1IdwPOt6verFmeePvg3eH8f/Mtkb7aamRVGR2bRpGTeVGmfpLHAqohYImlU2a6zgNERsUjSOcD3yJJ+zeTpotm62jFmZr1dF/bQjATGSRpNNm7ZIOkOYPeIWJSOuQW4u8K5K4GPlG03pjraqW9T1V9Zku7LU2dm1puVQrlLeyJiakQ0RsRQsgHTuWT95QPKBkwPJhuAbe12YKKkzSXtRNYl/iDZ8jC7StpJ0mbpurdXe6b21oPfAtgKGCxpENDyVA281ydkZlYItVxsLCLWSzoFmCGpBKwGTgSQNA4YHhEXRMRjkv6DbPB0PfC/IqI5HfcVYA7ZNMmfR8Rj1e6raGPoWNJk4GtkU3VW8l6CXwP8NCJ+tNFPm08BXzswsxrpdHae/6Ejcuec/V++tVesa9DeevA/AH4g6asR8cNujMnMrNtF539H9Dh5ho1Lkga2bEgalCbpm5kVxvpQ7tJb5Enwp0TEay0bEbEaOKV2IZmZdb9AuUtvkedFp36SFKmzPq2vsFltwzIz616legdQA3kS/N3ALZJ+krZPBe6qXUhmZt2vN7XM88qT4M8leyX3tLT9MPChPBeXtDlwODC0/F4RcXGHojQzq7EituCr9sFHRAlYRPYt1n3JPtdXaYJ+JbPIJvivB94sKxVJmiRpsaTFTU0V3wI2M6uJZpS79Bbtvei0G3B0Kq+QvVpLRHTkox+NEXFo3oNbre/gefBm1m0K+MW+drtongDmA2NbFqWX1NFP9f1O0l4R8cjGBmhm1h1Kvahlnld7Cf6LZOsd/FrS3cDN5HxbTNIjZC3w/sCXJT0DvJPOj4j4h05FbWbWxYrYZdDem6wzgZmSPkDWj/41YIikq4DbIuKedq47tmvDNDOrrSIOsuZZLvhN4EbgxrTo2L+QzaxpM8FHxHNdFqGZWTcoqW910WwgvcXa5kL3Zma9VXO9A6iBjflkn5lZ4fS1WTRmZn1GX5tFY2bWZ/SpWTRmZn2Ju2jMzAqqiNMk86wHb2ZWeM3KX/KQ1E/SMkmz0/Z8SctTeVHSzArnfLbsmOWS3pY0Ie27VtKzZfuGVYvBLXgzM2rSgp9MtjBjA0BE7N+yQ9IMssUY3ycifg0MS8dsAzzF+985Oicibs0bgFvwZmZkCT5vqUZSIzAGmFZhXwPZqrwbtOBbOQK4KyL+lvMRNuAEb2YGhPKX8qXNU5nU6nKXA1Oo/PtgAnBfRKypEtJE4KZWdd+U9LCk76fvbbTLCd7MjI614COiKSKGl5V33+6XNBZYFRFL2rjV0WyYuN9H0vbAXsCcsuqpwO7APsA2ZEvGtMsJ3syMbKmCvKWKkcA4SX8iW4X3QEnTASQNJvtw0h1VrnEk2aKO61oqIuKlyLwDXJOu0y4neDMzsnnweUt7ImJqRDRGxFCybpa5EXFs2n0EMDsi3q4Szgat/NSqR5LIunkerfZMTvBmZnTtIGs7NuhXlzRc0rSy7aHAR4DftDr3hvStjUeAwcAl1W6miB77gm6PDczMepxOv4f63R2PzZ1z/vW/pveK9149D97MjGK2KJ3gzczwWjRmZoXlD36YmRVUqYCdNE7wZmYUczVJJ3gzMzzIamZWWG7Bm5kV1HoVrw3vBG9mhrtozMwKy100ZmYF5WmSZmYFVbz07gRvZga4i8bMrLCaC9iGd4I3M8MteDOzwgq34M3MiskteDOzgiriNEl/k9XMjGyaZN6Sh6R+kpZJmp2250tansqLkma2cV5z2XG3l9XvJGmRpKck3SJps2oxOMGbmQHridwlp8nAipaNiNg/IoZFxDDgAeCXbZz3VstxETGurP5S4PsRsQuwGjipWgBO8GZmZIOsef+rRlIjMAaYVmFfA3AgULEF38b1lM65NVVdB0yodp4TvJkZ2SBr3iJpkqTFZWVSq8tdDkyh8tjtBOC+iFjTRihbpGsulNSSxLcFXouI9Wn7BWCHas/kQVYzMzo2TTIimoCmSvskjQVWRcQSSaMqHHI0FVr2ZT4aESsl7QzMlfQI8Hru4Mq4BW9mRsda8FWMBMZJ+hNwM3CgpOkAkgYD+wJ3tHVyRKxMfz4DzAM+CfwFGCippVHeCKysFogTvJkZ0ByRu7QnIqZGRGNEDAUmAnMj4ti0+whgdkS8XelcSYMkbZ5+Hkz2y+LxiAjg1+l8gOOBWdWeyQnezIxsHnze0gkTgZvKKyQNl9TSZfNxYLGk35Ml9G9FxONp37nA2ZKeIuuT/1m1mymq/DbqjDTyewywc0RcLGlH4EMR8WCO04v31oGZ1Yo6e4GjPzohd8656bmZnb5fd6h1C/5K4NNkgwoAbwA/buvg8pHppqaK4xdmZjXRhX3wPUatZ9HsFxGfkrQMICJWt/f2VauRabfgzazbFHGpglon+HWS+pGStaTt6F2/AM2sj/Bqkh13BXAbMETSN8lGgM+v8T3NzDqs2uyY3qimCT4ibpC0BDiIbBBkQkSsqHKamVm3cxfNRoiIJ4Anan0fM7POKGLfsZcqMDPDffBmZoXlLhozs4Kq5Uuf9eIEb2YGNLsFb2ZWTO6iMTMrKHfRmJkVlFvwZmYF5WmSZmYF5aUKzMwKyl00ZmYF5QRvZlZQRZxF42+ympnR9d9kldRP0jJJs9P2fEnLU3lR0swK5wyT9ICkxyQ9LOmosn3XSnq27BrDqsXgFryZGTWZRTMZWAE0AETE/i07JM0AZlU452/AcRHxpKQPA0skzYmI19L+cyLi1rwBuAVvZgY0Ryl3qUZSIzAGmFZhXwNwILBBCz4i/hgRT6afXwRWAdtt7DM5wZuZkfXB5y05XA5MofIy8xOA+yJiTXsXkLQvsBnwdFn1N1PXzfclbV4tCCd4MzM61gcvaZKkxWVlUst1JI0FVkXEkjZudTRwU3uxSNoe+HfgyxHv/pNhKrA7sA+wDXButWdyH7yZGR3rg4+IJqCpjd0jgXGSRgNbAA2SpkfEsZIGA/sCh7V17dSFcwfwvyNiYdk9X0o/viPpGuDr1eJ0C97MDChF5C7tiYipEdEYEUOBicDciDg27T4CmB0Rb1c6V9JmwG3A9a0HU1OrHkki6+Z5tNozOcGbmZG14PP+1wkTadU9I2m4pJbB2COBA4ATKkyHvEHSI8AjwGDgkmo3Uw+e3N9jAzOzHkedvcDuQ/bJnXOeWPVQp+/XHdwHb2YGVbteeiMneDMzvFywmVlhuQVvZlZQbsGbmRVUczTXO4Qu5wRvZkYxlwt2gjczwx/8MDMrLLfgzcwKyrNozMwKyrNozMwKKs+HPHobJ3gzM9wHb2ZWWO6DNzMrKLfgzcwKyvPgzcwKyi14M7OC8iwaM7OC8iCrmVlBFbGLxh/dNjOj6z+6LamfpGWSZqft+WUf0n5R0sw2zjte0pOpHF9Wv7ekRyQ9JekKSVW/C+sWvJkZNWnBTwZWAA3p+vu37JA0A5jV+gRJ2wAXAsOBAJZIuj0iVgNXAacAi4A7gUOBu9oLoGYteGU+Uqvrm5l1pVJE7lKNpEZgDDCtwr4G4ECgUgv+c8CvIuLVlNR/BRwqaXugISIWRvab6HpgQrU4apbgUxB3duQcSZMkLU7lVEDdXep1Xz+rn9fP26lnnUQnrV+7UnlLq1y1uML9LwemAJWm5kwA7ouINRX27QA8X7b9QqrbIf3cur5dte6DXyppn7wHR0RTRAxPpamWgbWj0/9H6UX60rOCn7fIuvVZW+Wq9+UrSWOBVRGxpI3TjwZu6o44a53g9wMekPS0pIfTAMHDNb6nmVk9jQTGSfoTcDNwoKTpAJIGA/sCd7Rx7kqgvGu7MdWtTD+3rm9XrQdZP1fj65uZ9SgRMRWYCiBpFPD1iDg27T4CmB0Rb7dx+hzg/0oalLYPAaZGxKuS1kgaQTbIehzww2qx1DTBR8Rztbx+jdSra6ge+tKzgp+3yHrLs04EvlVeIWk4cFpEnJwS+TeAh9LuiyPi1fTzGcC1wJZks2fanUEDoCJO7jczM7/oZGZWWE7wZmYF5QTfR0kaKOmMesfRHSQNlfRoveMw625O8H3XQLJBG7PCSG/QO68l/otIJJ0t6dFUvlbveLrBt4C/TwsfXVbvYLpBP0k/lfSYpHskbVnvgGpF0rGSHkz/2/5EUr96x1RL6V9of5B0PfAo759H3qc5wQOS9ga+TPZi1gjgFEmfrG9UNXce8HREDIuIc+odTDfYFfhxROwJvAYcXud4akLSx4GjgJERMQxoBo6pb1TdYlfgyojYs5dOz64JryaZ+Sfgtoh4E0DSL4H9gWV1jcq60rMRsTz9vAQYWsdYaukgYG/gobSa7JbAqrpG1D2ei4iF9Q6ip3GCt77inbKfm8kSXxEJuC69TdmXvFnvAHoid9Fk5gMTJG0l6QPAYamuyN4Atq53ENbl7gOOkDQEQNI2kj5a55isTpzggYhYSvYK8INk6zxMi4hCd89ExF+A+9Ogcl8YZO0TIuJx4HzgnrSw36+A7esbldWLlyowMysot+DNzArKCd7MrKCc4M3MCsoJ3sysoJzgzcwKygneakJSc1oL5VFJv5C0VSeuda2kI9LP0yTt0c6xoyR9ZiPu8af0vUyzwnCCt1p5K61z8wlgLXBa+U5JG/UWdfqs2ePtHDIK6HCCNysiJ3jrDvOBXVLrer6k24HHJfWTdJmkhyQ9LOlUeHfJ1x+lFQLvBYa0XEjSvPQNSyQdKmmppN9Luk/SULJfJGelfz3sL2k7STPSPR6SNDKdu21aVfIxSdPIXvE3KxSvRWM1lVrqnwfuTlWfAj4REc9KmgS8HhH7SNqc7M3ae4BPAh8D9gD+Dngc+Hmr624H/BQ4IF1rm/TB4quBv0bEd9JxNwLfj4gFknYk+2r9x4ELgQURcbGkMcBJNf2LMKsDJ3irlS0ltazeOB/4GVnXyYMR8WyqPwT4h5b+dWAA2bKvBwA3RUQz8KKkuRWuPwL4bcu1yr4839o/A3uklRUBGiR9MN3ji+ncOySt3sjnNOuxnOCtVt5K65G/KyXZ8lX/BHw1Iua0Om50F8axCTAiIt6uEItZobkP3uppDnC6pE0BJO2WVvP8LXBU6qPfHvhshXMXAgdI2imdu02qb71K5j3AV1s2JLX80vkt8D9T3eeBQV32VGY9hBO81dM0sv71pemj2D8h+1flbcCTad/1wAOtT4yIPwOTgF9K+j1wS9r1n8BhLYOswJnA8DSI+zjvzea5iOwXxGNkXTX/VaNnNKsbryZpZlZQbsGbmRWUE7yZWUE5wZuZFZQTvJlZQTnBm5kVlBO8mVlBOcGbmRXUfwPlFDrtcgwtJQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob = pipe1.predict_proba(X_test_ML)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob[:,1]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZIOYBfjnC48S",
        "outputId": "7a402050-f4d3-416c-e1df-0467ae507e30"
      },
      "execution_count": 234,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.4931972789115646\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob[:,1]) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "b39IGP4RDI8s",
        "outputId": "7983104f-755f-4b86-f236-f8db0b3bffcd"
      },
      "execution_count": 235,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Random Forest"
      ],
      "metadata": {
        "id": "0hRhX6FrMG3L"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pipe2 = Pipeline([\n",
        "                 (\"scale\", StandardScaler()),\n",
        "                 (\"classifier\", RandomForestClassifier(n_estimators = 100, random_state = 42))\n",
        "                 ])"
      ],
      "metadata": {
        "id": "KX2wmJx8DZrO"
      },
      "execution_count": 236,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pipe2.fit(X_for_ML,train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s9Ho7JfnDi5W",
        "outputId": "36a9b399-86b5-41a5-ed71-4088c79c34f4"
      },
      "execution_count": 237,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Pipeline(steps=[('scale', StandardScaler()),\n",
              "                ('classifier', RandomForestClassifier(random_state=42))])"
            ]
          },
          "metadata": {},
          "execution_count": 237
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Test set evaluation\n",
        "prediction = pipe2.predict(X_test_ML)\n",
        "# Visualization\n",
        "cmat = metrics.confusion_matrix(prediction,test_labels_encoded)\n",
        "s = sns.heatmap(cmat,annot=True,xticklabels=label, yticklabels=label)\n",
        "s.set(xlabel='Predicted', ylabel='Actual')\n",
        "acc = cmat.trace() / cmat.sum()\n",
        "print('Accuracy: {0:5.2f}%'.format(acc*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "cz3LJigzDnda",
        "outputId": "41a1d80a-e04e-482c-d068-1602420c88bc"
      },
      "execution_count": 238,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: 54.64%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob = pipe2.predict_proba(X_test_ML)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob[:,1]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UyA_sPyPEIT-",
        "outputId": "38d62cba-85fa-4613-ae2e-af7c1e872c1c"
      },
      "execution_count": 239,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.5504889455782312\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob[:,1]) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "ymjNTqTnEOrA",
        "outputId": "cd2d67bd-3dc3-43e3-ee39-93db6201d309"
      },
      "execution_count": 240,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## KNN"
      ],
      "metadata": {
        "id": "QxrLlQ3uMPr4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pipe3 = Pipeline([\n",
        "                 (\"scale\",StandardScaler()),\n",
        "                 (\"classifier\", KNeighborsClassifier(n_neighbors=7))\n",
        "                 ])"
      ],
      "metadata": {
        "id": "XBC_zxYbEj4M"
      },
      "execution_count": 241,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pipe3.fit(X_for_ML,train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fX78m5oMEeIF",
        "outputId": "97a276ab-f840-400d-bff8-65e24ae0fea2"
      },
      "execution_count": 242,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Pipeline(steps=[('scale', StandardScaler()),\n",
              "                ('classifier', KNeighborsClassifier(n_neighbors=7))])"
            ]
          },
          "metadata": {},
          "execution_count": 242
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob = pipe3.predict_proba(X_test_ML)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob[:,1]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oam2d5noF6kS",
        "outputId": "bc7f1d8d-d45c-49ae-bdd4-41126e2e3c7a"
      },
      "execution_count": 243,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.5491602891156462\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Test set evaluation\n",
        "prediction = pipe3.predict(X_test_ML)\n",
        "# Visualization\n",
        "cmat = metrics.confusion_matrix(prediction,test_labels_encoded)\n",
        "s = sns.heatmap(cmat,annot=True,xticklabels=label, yticklabels=label)\n",
        "s.set(xlabel='Predicted', ylabel='Actual')\n",
        "acc = cmat.trace() / cmat.sum()\n",
        "print('Accuracy: {0:5.2f}%'.format(acc*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "LAhKGaG4Eyoh",
        "outputId": "c713e93b-f79c-4cc7-c399-377aa247ff4a"
      },
      "execution_count": 244,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: 52.58%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob[:,1]) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "KeleiuOLE1Yb",
        "outputId": "80777d17-ecbd-4670-d289-fc3257b91867"
      },
      "execution_count": 245,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## AdaBoost"
      ],
      "metadata": {
        "id": "lWYVd2MoMW7f"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "clf = AdaBoostClassifier(n_estimators=200, random_state=0)\n",
        "clf.fit(X_for_ML,train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z8HWAHCWFBT6",
        "outputId": "6caba126-db7d-4a01-b4f6-ee7dd421b431"
      },
      "execution_count": 246,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "AdaBoostClassifier(n_estimators=200, random_state=0)"
            ]
          },
          "metadata": {},
          "execution_count": 246
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob = clf.predict_proba(X_test_ML)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob[:,1]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bahidginGSd7",
        "outputId": "f6113f88-ff75-400b-f6e4-020a2aaf331f"
      },
      "execution_count": 247,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.5189200680272109\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Test set evaluation\n",
        "prediction = clf.predict(X_test_ML)\n",
        "# Visualization\n",
        "cmat = confusion_matrix(prediction,test_labels_encoded)\n",
        "s = sns.heatmap(cmat,annot=True,xticklabels=label, yticklabels=label)\n",
        "s.set(xlabel='Predicted', ylabel='Actual')\n",
        "acc = cmat.trace() / cmat.sum()\n",
        "print('Accuracy: {0:5.2f}%'.format(acc*100))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "dHSeEpDxFOxA",
        "outputId": "c8bece85-4447-44e0-b2f5-81a6f1cb479a"
      },
      "execution_count": 248,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: 50.00%\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob[:,1]) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "JZh0f2SJFyC3",
        "outputId": "0db5a884-1d3c-4c3c-dcd1-b2e835998333"
      },
      "execution_count": 249,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## XGB"
      ],
      "metadata": {
        "id": "QfIx7G5NMdo7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "mod1 = xgb.XGBClassifier(booster='dart',learning_rate=0.1)\n",
        "train_arr = np.array(X_for_ML)\n",
        "test_arr = np.array(X_test_ML)\n",
        "mod1.fit(train_arr, train_labels_encoded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X6l_-aSAGbrJ",
        "outputId": "3d3ea8a3-84dd-4fef-c85a-3aabb364b5b3"
      },
      "execution_count": 250,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "XGBClassifier(booster='dart')"
            ]
          },
          "metadata": {},
          "execution_count": 250
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction = mod1.predict(test_arr)\n",
        "print(\"Accuracy = \", accuracy_score(test_labels_encoded, prediction))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "haEaOdTlH2YX",
        "outputId": "e6e785c4-ac07-4649-f28a-9b83dd3932a6"
      },
      "execution_count": 251,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy =  0.5927835051546392\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cm = confusion_matrix(test_labels_encoded, prediction)\n",
        "sns.heatmap(cm, annot=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "tN55xdyAISLT",
        "outputId": "87c5538c-d03e-4f0a-e8bb-d577c527baf1"
      },
      "execution_count": 252,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f2a95451d10>"
            ]
          },
          "metadata": {},
          "execution_count": 252
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob = mod1.predict_proba(test_arr)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob[:,1]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "g1IoFVCzIbcU",
        "outputId": "183e2d47-1c68-456a-9542-c2474d7b21b7"
      },
      "execution_count": 253,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.5787627551020409\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob[:,1]) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "NmoNZ-9MIjb7",
        "outputId": "a6832cf2-e83d-4587-ae0f-c95764964c2a"
      },
      "execution_count": 254,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXwV9dX48c9JQlbIRgIkEEgICQoIqJHFFdeildraWrVVq7Viad3bWlv7qLVPF1ttra0brVttUau1BVsffVqX+tSfIltAQUMCAbKwBMIWsifn98dM8BqyXEjmruf9et2X985878wZwHvmO9+Z8xVVxRhjTPSKCXYAxhhjgssSgTHGRDlLBMYYE+UsERhjTJSzRGCMMVHOEoExxkQ5SwTGGBPlLBGYiCMim0SkSUQaRGSbiDwpIkO7tTlRRF4Xkf0isldEXhKRSd3apIrI/SKyxd3WBvdzVmCPyBhvWSIwkWqeqg4FpgPHAt/rWiEis4H/BRYDuUABsBp4W0TGu23igdeAycBcIBWYDewCZngVtIjEebVtY3pjicBENFXdBryKkxC6/Bz4g6r+WlX3q2q9qv4AeBe4y21zBTAW+JyqrlPVTlXdoao/UtWXe9qXiEwWkX+KSL2IbBeR77vLnxSR//ZpN0dEqn0+bxKR74rIGuCA+/6Fbtv+tYg84L5PE5HHRGSriNSIyH+LSOwA/6hMFLNEYCKaiIwBzgUq3M/JwInA8z00/zNwtvv+LOAVVW3wcz/DgH8Br+D0Mibg9Cj8dSnwaSAdeBY4z90m7o/8F4FFbtsngXZ3H8cC5wBfO4x9GfMJlghMpPqbiOwHqoAdwJ3u8kycf/dbe/jOVqDr+v/wXtr05nxgm6rep6rNbk9j6WF8/wFVrVLVJlXdDKwEPueuOwNoVNV3RWQkcB5wk6oeUNUdwK+ASw5jX8Z8giUCE6k+q6rDgDnAUXz8A78b6ARyevhODrDTfb+rlza9yQM2HFGkjqpunxfh9BIAvsTHvYFxwBBgq4jsEZE9wKPAiAHs20Q5SwQmoqnqv3Eupdzrfj4AvANc1EPzL/Lx5Zx/AZ8SkRQ/d1UFjO9l3QEg2efzqJ5C7fb5eWCOe2nrc3ycCKqAFiBLVdPdV6qqTvYzTmMOYYnARIP7gbNFZJr7+TbgKyJyg4gME5EMdzB3NvBDt83TOD+6fxGRo0QkRkSGi8j3ReS8HvbxdyBHRG4SkQR3uzPddaU41/wzRWQUcFN/AatqHfAm8ARQqaofusu34tzxdJ97e2uMiBSKyGlH8OdiDGCJwEQB90f1D8Ad7uf/AJ8CLsQZB9iMM+h6sqqWu21acAaMPwL+CewD3sO5xHTItX9V3Y8z0DwP2AaUA6e7q5/GuT11E86P+HN+hr7IjWFRt+VXAPHAOpxLXS9weJexjPkEsYlpjDEmulmPwBhjopwlAmOMiXKWCIwxJspZIjDGmCgXdgWusrKyND8/P9hhGGNMWFmxYsVOVc3uaV3YJYL8/HyWL18e7DCMMSasiMjm3tbZpSFjjIlylgiMMSbKWSIwxpgoZ4nAGGOinCUCY4yJcp4lAhF5XER2iMgHvawXEXlARCpEZI2IHOdVLMYYY3rnZY/gSZxJv3tzLlDkvuYDD3sYizHGmF54lghU9S2gvo8mF+BMIK6q+i6QLiJWStcYY7rp6DhAU9Mmz7YfzDGC0Xxyer5qd9khRGS+iCwXkeV1dXUBCc4YY0LB7t2vs2zZVD7/4N+5+NF3PNlHWAwWq+pCVS1R1ZLs7B6fkDbGmIjS1raHsrJrWL36TERiSEoq9GxfwSwxUYMz4XeXMe4yY4yJaqodrFp1Io2NZeTl3Up+/l38ZqJ3+wtmIlgCXCcizwIzgb3ufKzGGBOV2tp2EReXiUgsBQU/JiEhj9TUEgBy073br2eJQESeAeYAWSJSDdwJDAFQ1UeAl4HzgAqgEbjKq1iMMSaUqSrbt/+JioobGT/+Z+TmXkN29uc+0eal1bUAzJuWO+j79ywRqOql/axX4Jte7d8YY8JBc3MV69d/nfr6l3m3bgGbYqZyWS7UH2hlwR9XHGy3bus+JuWkepIIwmKw2BhjItH27c+wbNlk9ux5kwkT7mdV/WU8/k5Tj20n5aRywfQeb6wcsLCbj8AYYyJFXFwGqakzKS5eSFJSAfAO2UMTAMhMiee5a2cHJo6A7MUYYwydne1UV/8K1VbGjbud4cPnkpn5KUQkqHFZIjDGmABoaFjNRx9dTUPDCrKzv4iqIiJBTwJgicAYYzyzaOkWFpdWcezI1czOuJG4uEwe+GAxQ4Zkw3/ePdju/Kk5XD47n5kFmSyt7KsyjzdssNgYYzyyuLSGdbV72b37DUaM+BIzZqxzkkAvFsyZ4NmAcF/EuYszfJSUlKhNXm+MCWXt7Q1c/vt/8cG2JCblpPLkFSNJShof1JhEZIWqlvS0znoExhgziOrr/8ny5cfQ0LCGo0bGccH00UFPAv2xMQJjjBmgRUu38LdVm2lq2kDZzmF8tvjTPP2100lPPyXYofnFegTGGDNAi0ur+aBmO62t2zh6RAtTi28OmyQA1iMwxhi/OXcBfVwkedqYBL533nQykhOYmZ/AA188h2HDwm/WXUsExhjjp8WlNQdr/rS2bmdr7Sts3Xocj1w+Hzg+2OEdMUsExhjjo/tZP8Ck3FTunDeZnLRERgzt4GtTfsju3a+SmnoiaWk3BSnSwWOJwBhjfLxfs4ellfXMLMg8ZN1tc9ZSXr6AvXuVCRN+w+jR30Ak/IdaLREYY4yPn144lZ9eOLXHdfHx2aSmnsTEiY+SmDguwJF5xxKBMcb0orOzjaqq+1BtIz//v8jM/BQZGeeERH2gwWSJwBgT0Xq65p+Tlsj9lxwLwA9fWsu62n0H1y2trOfSGXl8/+wOysqupqFhFSNGXBJSReIGW/hf3DLGmD5U7W48rEJuMwvSyUl4gxUrTqClpZbJk//CpEnPRGQC6GK1howxEaerF5CRHM8jlx/ebZ0NDR+wYsVxjBx5GYWF9zFkSIZHUQaW1RoyxkSVrvv9/dXe3sC2bU8DMHToFGbMKOOoox6PmCTQHxsjMMaEra4z/8QhsTz11RkAPPBa+cGHvvzpDdTXv0pZ2XxaWqoYNqyElJSj3Wkjo4f1CIwxYau3M39/Jnpva9vFhx9+hTVr5hIbm8yxx/4fKSlHexVqSLMegTEmJHSd3T951QyS4mN5+p1N/H3N1kPadU3ovvCtDQfP/Lt6AwA3nFnEDWcW9bkv1Q5WrjyJpqYKxo69nXHjfkBsbOKgHk84sURgjAkJi0trDnuaRn/O/H21ttYxZMhwRGIpLLyHhIRxDBs2/XBDjTh215AxJiRc/Og7wMdn/INJVdm27Uk2bLiF8eN/Rm7utYO+j1DX111D1iMwxkS0pqZNrF8/n927/0la2imkp58e7JBCjg0WG2OCZtHSLdQfaAWgrqFl0Le/bdvTLFs2hX373qGo6CGmT3+T5OTiQd9PuLMegTEmaBaX1lC5s4HbPz2JBacV0tYxuJeq4+NHkp5+KsXFj5CYOHZQtx1JLBEYYzy3aOkW5kzMJjc9iZdW1/LHdzcDfOLWz4tK8ga8H6dI3M9R7SA//w4yM88hM/OcAW830tmlIWOM5xaX1vDwmxsOWX64d/30Zf/+laxYcQKVlT+gsbGMcLsRJpisR2CM8YRv1U/fM/9503KZNy130PbT0dHEpk0/pKrqXuLjs5k8+a9kZ3920LYfDTztEYjIXBEpE5EKEbmth/VjReQNEVklImtE5Dwv4zHGBM6IYQm0dzpn5YN55t9dc/NGqqt/yahRV3LCCessCRwBz54jEJFYYD1wNlANLAMuVdV1Pm0WAqtU9WERmQS8rKr5fW3XniMwxrS376Ou7kVycq4EoLl5c0TNGOaFYFUfnQFUqOpGVW0FngUu6NZGgVT3fRpQ62E8xpgA2lDXwIa6hkHf7q5dL7Ns2RTKyq7mwIEPASwJDJCXiWA0UOXzudpd5usu4DIRqQZeBq7vaUMiMl9ElovI8rq6Oi9iNcYMsu+/+D7ff/H9Qdtea+tOPvzwct5//9PExg7j2GPfjtoicYMt2HcNXQo8qapjgPOAp0XkkJhUdaGqlqhqSXZ2dsCDNMYEl2oHq1adxI4dzzJu3B2UlKwkLW1WsMOKGF7eNVQD+N4YPMZd5utqYC6Aqr4jIolAFrDDw7iMMWGitXU7Q4Zku0Xi7iUxcRxDh04NdlgRx8sewTKgSEQKRCQeuARY0q3NFuBMABE5GkgE7NqPMVFOVdm69TGWLp1Ibe1CALKy5lkS8IhnPQJVbReR64BXgVjgcVVdKyJ3A8tVdQnwLeB3InIzzsDxlWpPgRgT1ZqaNlJWdg179rxOWtppZGScFeyQIp6nD5Sp6ss4g8C+y+7web8OOMnLGIwxwXH9GX1PDtOTbdueYv36byASS3HxI+TkXEMPw4ZmkNmTxcYYT5xclHXY34mPzyUj4wyKih4mMXGMB1GZnlgiMMZ4Ym3tXgAm56b12qazs5UtW36GaicFBXeRmXk2mZlnBypE47JEYIzxxN0vOUUEeptxbN++ZZSVfZUDBz5g5MjLUVVEJJAhGpclAmOM39bW7j34A+/r1rkTOX5cJis21/PzV8oADk4s311HRyOVlXdQXf0r4uNzmDJlCVlZ8zyP3fTOEoExpl//Kd8JQEbKEL+/01uhuebmSmpqfkNOzjUUFt5DXFzvl45MYNjk9caYQ2yoa/hEeYius/sjnVi+vX2vWyTuKgCam6tITBz4RDTGfzZ5vTHGL/9atx2AguyUTywfSBnpXbv+QVnZtbS2biU1dTYpKUdZEggxlgiMiXLdJ5DpOvM/0rP/Lq2tdVRU3MSOHYtISZnClCkvkpJy1GCEbAaZJQJjotzi0pqDCWCwJpBxisSdTHNzJfn5P2Ts2NuIiYkfhGiNFywRGBNifM/Qzzx6BPNPLQTg4kffOaTt+VNzuHx2Pk2tHVz5xHuHrP/C8WO4qCSP+gOtLPjjikPWXzZrHL+6eDoAuelJA469pWUb8fEj3CJx95GYmM/QoVMGvF3jLXt225gQ03WGHii56UkDTgKqndTWPsp77xVTW/soAFlZ51sSCBN+3TUkIknAWFUt8z6kvtldQyZS+J75nzQhixvOdGrz5N/2D2YWZA74Gn2gNDZWsH79NezZ8ybp6WcwceLvSEoaH+ywTDcDmqpSROYBpcAr7ufpItK9nLQx5jD1duZ/WnG2ZxO9D7atW59g+fJj2L9/JcXFv2PatH9ZEghD/owR3IUz//CbAKpaKiIFHsZkTFjwPaPvMik3lTvnTQbgpmdXsXVv8yfWHzcug+/Ode6cWVpZ3+OZ/1NfneFh1IMrMXEsGRmforj4QRISwiN5mUP5kwjaVHVvtxog4fUUmjEeeL9mz8Ef8yMxd/IoTi0Or6lXOztb2Lz5p0AnBQV3k5FxJhkZZwY7LDNA/iSCtSLyJSBWRIqAG4D/521YxoSOns78x2en8NMLp/LTC3ufMev+S47tc7uPXH78oMQXKPv2LeWjj66msXEtI0d+xYrERRB/7hq6HpgMtACLgL3AjV4GZUwoKd+xn6WV9cEOI2g6Og5QUXELK1fOpqNjL8cc83eOPvpJSwIRpN+7hkTkIlV9vr9lgWJ3DRkTWAcOrGP58uPIyfkq48f/jLi4QyuKmtA3oLuGgO/5ucwYEyHa2vZQW/t7AFJSJjFzZgXFxQ9ZEohQvY4RiMi5wHnAaBF5wGdVKtDudWDGhIqbnl0F9H/NP1Ls3LmY9esX0Nq6g7S0k90icTZtZCTra7C4FlgOfAbwfTZ9P3Czl0EZE0q63wIaqVpbd1BefgN1dc+RkjKVKVOWWJG4KNFrIlDV1cBqEVmkqm0BjMkYE2BOkbiTaG7eQkHBf5OXdysxMf5PQmPCmz+3j+aLyE+BSUBi10JVtccHjQlzLS21xMePQiSWCRN+TWJiPikpk4IdlgkwfwaLnwAexhkXOB34A/BHL4MyxnhLtZOamod5772jqK19BIDhw8+zJBCl/OkRJKnqayIiqroZuEtEVgB3eBybMSHhuHEZwQ5hUDU2rqes7Br27n2LjIyzyMw8N9ghmSDzJxG0iEgMUC4i1wE1wFBvwzImuLqeJs5Ijg+7J4D7snXrY5SXX0dMTCITJz7OqFFX2oNhxq9LQzcCyTilJY4HLgO+4mVQxgRboOcECJTExHwyM8/lhBPWkZNzlSUBA/TTIxCRWOBiVf020ABcFZCojPFQT7WDEofEHqz6+cBr5Qenbgz33kBnZwubNv0IgPHj/9uKxJke9dkjUNUO4OQAxWJMQDS0tLF++/4+2wzW3L3BtHfv/2P58uls2fJjWlu34s8kVCY6+VNr6GFgNPA8cKBruaq+6G1oPbNaQ8b0rb29gcrK26mp+Q0JCXlMnLiQzMxPBTssE2QDrTWUCOwCzgDmua/z/dzxXBEpE5EKEbmtlzZfFJF1IrJWRBb5s11jTO9aWrZQW/soo0d/kxNO+MCSgOlXv3cNqeoRjQu44wsPAmcD1cAyEVmiqut82hThFLA7SVV3i8iII9mXMf3xHRfobWawcNbWtpu6uufJzZ1PSsokZs3aSEJCbrDDMmHCnx7BkZoBVKjqRlVtBZ4FLujW5hrgQVXdDaCqOzyMx0Qx37uAZhZkhv31f191dX9l2bJJrF//DRobywAsCZjD4s9zBEdqNFDl87kamNmtTTGAiLwNxAJ3qeor3TckIvOB+QBjx471JFgT2c6fmsP5U3O4fHZ+sEMZNC0t26iouJ66uhcYOnQ6xxzzD5KTJwY7LBOGvEwE/u6/CJgDjAHeEpFjVHWPbyNVXQgsBGewONBBmvAXSQkAnCJxpaWn0NxcRUHBT8jL+7YViTNHrN9EICIjgZ8Auap6rohMAmar6mP9fLUGyPP5PMZd5qsaWOpWN60UkfU4iWGZvwdgjD+aWjsASIqPDXIkA9PcXE1CQq5bJO4BEhMLrFS0GTB/xgieBF4Fui46rgdu8uN7y4AiESkQkXjgEmBJtzZ/w+kNICJZOJeKNvqxbWMOy5VPvMeVT7wX7DCOmGon1dW/cYvEPQzA8OHnWhIwg8KfRJClqn8GOgFUtR3o6O9LbrvrcJLIh8CfVXWtiNwtIp9xm70K7BKRdcAbwHdUddcRHIcxEevAgY9YtepUKipuIC3tZIYP9+vubWP85s8YwQERGQ4ogIjMAvb6s3FVfRl4uduyO3zeK3CL+zJmQHxvEX34suPJTInn+eVVvLCi+mDJiHBTW/t7ysuvIzY2maOOeoqRIy+3+kBm0PmTCL6Fc0mn0L27Jxv4gqdRGXMEum4R7ekHP1xLRiQlFZKVNY+iot8SHz8y2OGYCNVviQkAEYkDJgIClAVz6korMWF6c/Gj7wCE9YNiHR3NbN58NwDjx/8kyNGYSDKgEhMisga4FWhW1Q9s/mJjvLF379tukbif0tZWZ0XiTMD4c2loHnAx8GcR6QSewxn43eJpZMYcpstmjQt2CEekvX0/lZXfp6bmQRITxzF16qtkZp4T7LBMFPHr0tDBxk5toP8CvqyqQbkh2y4NmUhz4MCHrFhxHDk58yko+DFxcTYBoBl8fV0a8uvJYhEZh9MruBjn1tFbBy88YwZH7Z4mAHLTk4IcSf/a2naxY8efGT16ASkpRzNz5kYSEnKCHZaJUv48WbwUGIIzH8FFqmoPfJmQdPNzpUBoDxarKnV1f6G8/Ju0t9eTkXEGyckTLQmYoPKnR3CFqpZ5HokxEa6lZSvl5d9k586/MnTo8Uyb9r9WJM6EhF4TgYhcpqp/BD4tIp/uvl5Vf+lpZMZEENUOVq06hdbWGsaP/zljxtxMTEywaz4a4+jrX2KK+99hPayz+9pMSPnXuu0h+fRwc3MVCQmjEYmluPhBEhMLSE4uDnZYxnxCr4lAVR913/5LVd/2XSciJ3kalTF+2lDXcPB9KD09rNpBTc2DbNz4PQoLf87o0d+0KSNNyPKnb/ob4Dg/lhnjuf+U7+Q3r5cf/NzVC3ju2tmcNSk0SjAcOPAhZWVXs2/fO2Rmnsvw4fOCHZIxfeprjGA2cCKQLSK+ReFScWYTMyZg1tb2XOcwlHoBALW1Cykvv57Y2GEcddTTjBz5ZSsSZ0JeXz2CeGCo28Z3nGAfVnTOBIBvNVHfM/+Ti7KCHFnvkpKKyMr6HEVFDxAfPyLY4Rjjl77GCP4N/FtEnlTVzQGMyRjgk9VEQ+3Mv0tHRxObNt0FCIWFPyMj43QyMk4PdljGHJa+Lg3dr6o3Ab8VkUPuElLVz/TwNWMGza1znXvsjx+XGeRIerZnz1uUlX2NpqZycnO/jqraZSATlvq6NPS0+997AxGIMd2FagJob9/Hxo23UVv7MImJ45k27TUyMs4IdljGHLG+Lg2tcP/7765lIpIB5KnqmgDEZqKQ77jAsMQ4FswpDLmE0NJSy7ZtTzJmzC0UFNxNbGxK/18yJoT5Mx/BmyKSKiKZwErgdyJiTxUbT3SNCwDsb26nbFtDP98IjNbWndTUPARASspRzJpVyYQJ91kSMBHBn+cI0lR1n4h8DfiDqt7pTlZjjCe67g4KBU6RuD9TXn497e17yMg4i+TkYps20kQUfxJBnIjkAF8Ebvc4HhPl7pg3KdghHNTSUsv69QvYtWsJw4aVMHHia1YewkQkfxLB3cCrwNuqukxExgPl/XzHmD75jgX4uv6MopB4TsApEncqra01FBbey+jRN1qROBOx+v2XrarP48xF0PV5I/B5L4MykW/m+ExWbtlNVX1jsEP5hObmzSQkjHGLxD1EYuJ4kpMnBDssYzzlz8Q0Y3BqC3UVmvs/4EZVrfYyMBMZavc0HZwwxtc1p4zn3oumBSGinql2UF39ayorf+CWib7O5g02UaPfu4aAJ4AlQK77esldZkxEaGj4gJUrT2TDhm+RkXEmWVmfDXZIxgSUPxc9s1XV94f/SRG5yauATHjzvfa/v7mdBXMKQ+YOoJ7U1DxCRcUNxMWlcfTRixgx4hJ7OthEHX96BLtE5DIRiXVflwG7vA7MhCff5wCGJcaxv7k9yBH1TNWpmpKScjTZ2RdxwgnrGDnyUksCJir50yP4Ks4Ywa/cz28DV3kWkQl7ofQcQHcdHY1UVt6BSCyFhfeQnn4a6emnBTssY4LKn7uGNgNWYM745eHLjg92CL3avftNysq+RnPzBnJzv2FF4oxx+VNiYryIvCQidSKyQ0QWu88SGHOIzJR4MlPigx3GJ7S376Ws7FpWr3bKQ0+b9jrFxQ9aEjDG5c8YwSLgz0AOzl1DzwPPeBmUCV/PL6/i+eVVwQ7jE1patrJ9+x/Jy/s2J5ywxuYLMKYbfxJBsqo+rart7uuPQKI/GxeRuSJSJiIVInJbH+0+LyIqIiX+Bm5C0wsrqnlhRfAfMWltraO6+jdAV5G4TRQW/oLY2OQgR2ZM6PFnsPh/3B/xZwEFLgZedquRoqr1PX1JRGKBB4GzgWpgmYgsUdV13doNA24Elh7xUZiAW7R0Cx2dnVw+O5+m1g6ufOI94OMpJYNFVdmx4xnKy2+go2MfmZmfcovEZQctJmNCnT+J4Ivuf6/ttvwSnMTQ23jBDKDCLUmBiDwLXACs69buR8A9wHf8CdiEhsWlNVTVN3L57PxPLA/mlJLNzVWsX7+A+vp/MGzYTI466jErEmeMH/y5a6jgCLc9GvC9WFwNzPRtICLH4Ux08w8R6TURiMh8YD7A2LFjjzAcM9jyMp3LLEnxsUG/XbSzs53S0jm0tm6jsPBXjBlzPU6n1BjTn6CVUxSRGOCXwJX9tVXVhcBCgJKSkkPmTzbRq6lpE4mJecTExFFc/ChJSeNJSrKb2ow5HP4MFh+pGiDP5/MYd1mXYcAU4E0R2QTMApbYgLHxR2dnO1u23MuyZUcfnDksM/MsSwLGHAEvewTLgCIRKcBJAJcAX+paqap7gYOF50XkTeDbqrrcw5hMBGhoWENZ2dXs37+c4cMvIDvbqqIbMxD+lKEW4MvAeFW9W0TGAqNU9b2+vqeq7SJyHc6kNrHA46q6VkTuBpar6pJBiN8ESbDGBGpqHqKi4kbi4jKYNOk5srMvsgfDjBkgf3oEDwGdwBk4s5XtB/4CnNDfF1X1ZeDlbsvu6KXtHD9iMVGqqxxESsoURoy4hMLCXxEfH/yZzIyJBP4kgpmqepyIrAJQ1d0iElo1BEzALXxrAwDzTy30dD8dHQeorPwBInEUFv6C9PRTSU8/1dN9GhNt/BksbnMfDlMAEcnG6SGYKPbahzt47cMdnu5j9+7XWLbsGKqr76ezs+Vg6WhjzODyp0fwAPBXYISI/Bj4AvADT6MyUa2tbQ8bNnybbdseIympiOnT3yI9/ZRgh2VMxPLngbI/icgK4ExAgM+q6oeeR2aiVlvbdnbseJa8vO+Sn38nsbFJwQ7JmIjmz11DY4FGnLmKDy5T1S1eBmaiS2ur8+M/ZsyNJCdPZNasTTYYbEyA+HNp6B844wOCU3W0ACgDJnsYlwlxiUMGp3yDqrJ9+5+oqLiRjo4GMjPPIzm5yJKAMQHkz6WhY3w/u/WBvuFZRCak+E5G3+WkCVk89dUZA952c/MW1q//OvX1/0Nq6mwmTnyM5OSiAW/XGHN4DvvJYlVdKSIz+29pIsFb6+tYWlnPzILMQd3ux0XidjBhwgOMHv0NKxJnTJD4M0Zwi8/HGOA4oNaziExIeeTywZ2DuKlpI4mJ44iJiWPixN+RmFhIUlL+oO7DGHN4/HmOYJjPKwFnzOACL4MykccpEncP7703iZqaBwHIyDjTkoAxIaDPHoH7INkwVf12gOIxIcB3XGBpZT0L5hTy3blHHfH29u8vpazsahoaVpKV9Tmysy8arFCNMYOg1x6BiMSpagdwUgDjMSFgcWkN67buA2BmQSZ5GUc+z2919W9ZufIEWlpqmDz5BaZMeZGEhJzBCtUYMwj66hG8hzMeUCoiS4DngQNdK1X1RY9jMwG0aOkWynfs5855k8lJSyQnLZH7Lzn2iLfXVSRu6NCpjBjxZSZM+CVDhgzugLMxZnD4c1joYnMAABLOSURBVNdQIrALp/po1/MEClgiiCCLS2tYWlnPnfMmDygBtLc3UFl5OyJDmDDhXisSZ0wY6CsRjHDvGPqAjxNAF6v+FaYWLd3C+zV7+OmFUwH43otr2Fh3gHVb9w34FtH6+v+lrGw+LS1bGD36+oO9AmNMaOvrrqFYYKj7GubzvutlwtDi0hqeea/qkOWTclK5YProI9pmW9tuPvroKtas+RQxMYlMn/4WRUW/tiRgTJjoq0ewVVXvDlgkJmB8z/y7egYD0da2g7q6Fxg79nuMG3cHsbGJA96mMSZw+koEdjpnetXSso0dO54hL+/mg0XihgwZHuywjDFHoK9LQ2cGLArjmUVLt3DTs6sOfl5aWT+g7akq27Y9xbJlk9i48Xs0NpYDWBIwJoz1mghUdWC/GCYkLC6t4W+lH1cEueqk/CMeC2hq2sSaNXP56KMrSU6eRElJqRWJMyYCHHbROROaFi3dwlvr6w7WBrrnlY9YuXn3IXcD3TnvyKqHd3a2s3r16bS17aSo6EFyc7+OiD8VSowxoc4SQYToeg6gu4HcDQTQ2FhBUlKBWyTucZKSxpOYOG4goRpjQowlggjie+Y/kNpAAJ2dbVRV/YJNm35IYeEvGDPmBjIyTh9oiMaYEGSJwBxi//6VbpG4UrKzL2LEiIuDHZIxxkOWCMwnVFc/QEXFLcTHZzN58otkZ38u2CEZYzxmiSBCnDRhYHP8flwk7lhGjbqCwsL7GDIkY5CiM8aEMksEEeKGM4/sNs729v1s3Pg9YmISmDDhPtLTTyE9/ZRBjs4YE8rs/r8otmvXKyxbNoXa2ocARdVqCRoTjaxHECG+8vh7ADz11Rn9tm1r20VFxS1s3/4HkpOP5thj3yYtbbbXIRpjQpQlgjDgO3UkwHPXOj/aC9/awGsf7gBg3dZ9TMpJ9Wt7bW272Lnzr4wb91+MG3c7MTEJgx+0MSZseHppSETmikiZiFSIyG09rL9FRNaJyBoReU1E7EmlHvhOHdmb/h4ca2nZypYt96KqJCcXM2vWZgoK7rYkYIzxrkfgTnz/IHA2UA0sE5ElqrrOp9kqoERVG0VkAfBzIGpvWu9+5v/kVTNIio+lqr6RSTmpB3sCXeafWsj8Uwv73KZTJO4JKipuQbWFrKwLSE4usjuCjDEHedkjmAFUqOpGVW0FngUu8G2gqm+oaqP78V1gjIfxhLzezvwXzCk8ojIRTU2VrFlzDmVlVzN06DRKSlZbkThjzCG8HCMYDfhOhVUNzOyj/dXA//S0QkTmA/MBxo4dO1jxhYRFS7cwd8ooMlPiqWto6fHM//LZ+Ye9XadI3Bm0te2iqOhhcnPnW5E4Y0yPQmKwWEQuA0qA03par6oLgYUAJSUlEXWP4+LSGip3NnD7pyex4LRC2joGdniNjeUkJY13i8Q9QVJSIYmJeYMUrTEmEnmZCGoA31+gMe6yTxCRs4DbgdNUtcXDeELWmuq9AFxUcuQ/2J2dbWzZcg+bN/+IwsKfM2bMjWRkzBmkCI0xkczLRLAMKBKRApwEcAnwJd8GInIs8CgwV1V3eBhLRNu3bzllZVdz4MAaRoy4hBEjLg12SMaYMOJZIlDVdhG5DngViAUeV9W1InI3sFxVlwC/AIYCz4sIwBZV/YxXMUWi6upfu0XiRjFlymKysuyPzxhzeDwdI1DVl4GXuy27w+f9WV7uP5J1FYkbNqyEnJyrGT/+5wwZkh7ssIwxYSgkBouj2WWzDu8Zuvb2fWzc+F1iYhKZMOFXpKWdRFraSR5FZ4yJBpYIgmzetFy/2+7a9TLr119LS0steXm3HOwVGGPMQFgiCLLaPU0A5KYn9dqmtXUnFRU3sWPHn0hOnsxxx71Aampfj2QYY4z/LBEE2c3PlQIc8hCZr/b23eza9RLjxt3JuHHfJyYmPlDhGWOigCWCENXSUsP27X8iL+87JCcXMWvWZhsMNsZ4whJBiFFVtm79PRs2fBvVNrKyLiQ5eYIlAWOMZywRhJCmpg2UlV3Dnj1vkJ4+h+Li35GcPCHYYRljIpwlghDR2dlOaemZtLfXU1z8KDk5X7MiccaYgLBEEGRXzEggPn4UMTFxHH30UyQmFpKYGNXVuI0xAWannEHS2dnKpk0/ZOi+WRyd9jwA6emnWRIwxgSc9QiCYN++99wicR/QmnAtDTEXBjskY0wUsx5BgFVV3c/KlbNpa9vNlCkv8diar3DnS9XBDssYE8UsEQSIqjPhTGrqDHJyrmHGjLVkZZ0f5KiMMcYuDXmuvX0vGzbcSkxMEkVF95OWdiJpaScGOyxjjDnIEoGfFi3dwuJSZ4K1688o4uSiLNbW7uXul9Yd0vbWuRM5flwmr61+iXv/t5zOzgISE/NITHwHgDvmTWJybhr/Kd/Juq37mJSTGtBjMcYYX3ZpyE+LS2tYt3WfX23b2vawbt2X2LjxVkSGMGzYcSQmju+x7aScVC6YPnowQzXGmMMiXdeuw0VJSYkuX7484Pu9+FHnbL6v4nBdGhsrWLGihLy8Wxg79jYrEmeMCToRWaGqJT2ts0tDg6S5uYrt2//I2LG3kZw8gdmzNxMXlxbssIwxpl+WCPz0kwuP6XG5aie1tQvZuPFWVDvIzr6I5OQJlgSMMWHDEkE3voPCvq45ZTxnTRr5iWWNjeWUlV3D3r3/Jj39TCZOXEhSUs9jAcYYE6osEXQzZ2I2H27dx/rt+/ts19nZzurVZ9PevoeJEx9j1KirbNpIY0xYskTQTW56Ej/67JRe1x848CFJSUVukbinSUoqJCHB/3mHjTEm1Njto928tLqWl1bXHrK8s7OFyso7Wb58KjU1vwUgPf0USwLGmLBnPYJu/vjuZgDmTfv4B37v3ncpK7uaxsZ1jBx5OaNGXR6s8IwxZtBZIuhHVdV9bNjwHRISxnDMMS8zfPi5wQ7JGGMGlV0actUfaOXiR985+PSwaicAqamzyc39Oiec8IElAWNMRLJE0M3Ro5KZOepfVFTcCEBa2okUFz9EXJzVAzLGRKaovDTU/VmBuoYWFpxWyG8v3E55+Tdobd1BbOytqKrdEmqMiXhR2SNYXFrD0sr6g5+zUmLYXPUIa9d+jvj4kRx//HuMH/8TSwLGmKgQNT2Crl7A+VNzePKqGQAkxccCTpG4lSsfIi/vx+TlfYeYmCHBDNUYYwIqanoEvmWkk+Jjkc4aNm/+MapKcvIEZs3awrhx37ckYIyJOp4mAhGZKyJlIlIhIrf1sD5BRJ5z1y8VkXwv45mUk8pls8ZSU/MQy5ZNZvPmn9DUtAGAuLhhXu7aGGNClmeJQERigQeBc4FJwKUiMqlbs6uB3ao6AfgVcI9X8QB0djZSWjqH8vJvkpo6mxNOWEty8gQvd2mMMSHPyx7BDKBCVTeqaivwLHBBtzYXAE+5718AzhTPRmiVhoY1HDjwPhMnPsHUqa+SlJTvza6MMSaMeDlYPBqo8vlcDczsrY2qtovIXmA4sNO3kYjMB+YDjB079oiCee7aE9mzp5OkpHUkJOQc0TaMMSYShcVgsaouVNUSVS3Jzs4+4u2kp59sScAYY7rxMhHUAHk+n8e4y3psIyJxQBqwy8OYjDHGdONlIlgGFIlIgYjEA5cAS7q1WQJ8xX3/BeB1VVUPYzLGGNONZ2ME7jX/64BXgVjgcVVdKyJ3A8tVdQnwGPC0iFQA9TjJwhhjTAB5+mSxqr4MvNxt2R0+75uBi7yMwRhjTN/CYrDYGGOMdywRGGNMlLNEYIwxUc4SgTHGRDkJt7s1RaQO2HyEX8+i21PLUcCOOTrYMUeHgRzzOFXt8YncsEsEAyEiy1W1JNhxBJIdc3SwY44OXh2zXRoyxpgoZ4nAGGOiXLQlgoXBDiAI7Jijgx1zdPDkmKNqjMAYY8yhoq1HYIwxphtLBMYYE+UiMhGIyFwRKRORChG5rYf1CSLynLt+qYjkBz7KweXHMd8iIutEZI2IvCYi44IR52Dq75h92n1eRFREwv5WQ3+OWUS+6P5drxWRRYGOcbD58W97rIi8ISKr3H/f5wUjzsEiIo+LyA4R+aCX9SIiD7h/HmtE5LgB71RVI+qFU/J6AzAeiAdWA5O6tfkG8Ij7/hLguWDHHYBjPh1Idt8viIZjdtsNA94C3gVKgh13AP6ei4BVQIb7eUSw4w7AMS8EFrjvJwGbgh33AI/5VOA44INe1p8H/A8gwCxg6UD3GYk9ghlAhapuVNVW4Fnggm5tLgCect+/AJwpIhLAGAdbv8esqm+oaqP78V2cGePCmT9/zwA/Au4BmgMZnEf8OeZrgAdVdTeAqu4IcIyDzZ9jViDVfZ8G1AYwvkGnqm/hzM/SmwuAP6jjXSBdRAY0B28kJoLRQJXP52p3WY9tVLUd2AsMD0h03vDnmH1djXNGEc76PWa3y5ynqv8IZGAe8ufvuRgoFpG3ReRdEZkbsOi84c8x3wVcJiLVOPOfXB+Y0ILmcP9/75enE9OY0CMilwElwGnBjsVLIhID/BK4MsihBFoczuWhOTi9vrdE5BhV3RPUqLx1KfCkqt4nIrNxZj2coqqdwQ4sXERij6AGyPP5PMZd1mMbEYnD6U7uCkh03vDnmBGRs4Dbgc+oakuAYvNKf8c8DJgCvCkim3CupS4J8wFjf/6eq4ElqtqmqpXAepzEEK78OeargT8DqOo7QCJOcbZI5df/74cjEhPBMqBIRApEJB5nMHhJtzZLgK+4778AvK7uKEyY6veYReRY4FGcJBDu142hn2NW1b2qmqWq+aqajzMu8hlVXR6ccAeFP/+2/4bTG0BEsnAuFW0MZJCDzJ9j3gKcCSAiR+MkgrqARhlYS4Ar3LuHZgF7VXXrQDYYcZeGVLVdRK4DXsW54+BxVV0rIncDy1V1CfAYTvexAmdQ5pLgRTxwfh7zL4ChwPPuuPgWVf1M0IIeID+POaL4ecyvAueIyDqgA/iOqoZtb9fPY/4W8DsRuRln4PjKcD6xE5FncJJ5ljvucScwBEBVH8EZBzkPqAAagasGvM8w/vMyxhgzCCLx0pAxxpjDYInAGGOinCUCY4yJcpYIjDEmylkiMMaYKGeJwIQsEekQkVKfV34fbRsCF1nvRCRXRF5w30/3rYQpIp/pq0qqB7Hki8iXArU/E77s9lETskSkQVWHDnbbQBGRK3Eqnl7n4T7i3HpZPa2bA3xbVc/3av8mMliPwIQNERnqzqWwUkTeF5FDqo2KSI6IvOX2ID4QkVPc5eeIyDvud58XkUOShoi8KSK/9vnuDHd5poj8za39/q6ITHWXn+bTW1klIsPcs/AP3Kdg7wYudtdfLCJXishvRSRNRDa79ZAQkRQRqRKRISJSKCKviMgKEfk/ETmqhzjvEpGnReRtnAcj8922K93XiW7TnwGnuPu/WURiReQXIrLMPZZrB+mvxoS7YNfetpe9envhPBlb6r7+ivMkfKq7LgvnycquXm2D+99vAbe772Nxag5l4cxJkOIu/y5wRw/7exP4nfv+VNx68MBvgDvd92cApe77l4CT3PdD3fjyfb53JfBbn+0f/AwsBk53318M/N59/xpQ5L6fiVP+pHucdwErgCT3czKQ6L4vwnniFpynU//u8735wA/c9wnAcqAg2H/P9gr+K+JKTJiI0qSq07s+iMgQ4CcicirQiVN6dySwzec7y4DH3bZ/U9VSETkNZ8KSt93yGvHAO73s8xlwasKLSKqIpAMnA593l78uIsNFJBV4G/iliPwJeFFVq8X/aS2ew0kAb+CUOHnI7aWcyMdlQMD5we7JElVtct8PAX4rItNxkmdxL985B5gqIl9wP6fhJI5Kf4M2kckSgQknXwaygeNVtU2cqqKJvg3cH/BTgU8DT4rIL4HdwD9V9VI/9tF90KzXQTRV/ZmI/AOn7svbIvIp/J8AZwlOUssEjgdeB1KAPb7Jrw8HfN7fDGwHpuFc7u0tBgGuV9VX/YzRRAkbIzDhJA3Y4SaB04FD5l0WZy7m7ar6O+D3OFP+vQucJCIT3DYpItLbWfPFbpuTcao67gX+DycJdQ3A7lTVfSJSqKrvq+o9OD2R7tfz9+NcmjqEqja43/k1zuWbDlXdB1SKyEXuvkREpvn557JVnfr7l+NcEutp/68CC9zeEiJSLCIpfmzfRDjrEZhw8ifgJRF5H+f69kc9tJkDfEdE2oAG4ApVrXPv4HlGRLoutfwAp1Z/d80isgrncstX3WV34VxuWoNT7bGrhPlNbkLqBNbizPrmO2XgG8BtIlIK/LSHfT0HPO/G3OXLwMMi8gM3hmdx5unty0PAX0TkCuAVPu4trAE6RGQ18CRO0skHVopz7akO+Gw/2zZRwG4fNcYlIm/i3G4ZznMWGHPY7NKQMcZEOesRGGNMlLMegTHGRDlLBMYYE+UsERhjTJSzRGCMMVHOEoExxkS5/w8w+9w/YyJHIgAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## LightGBM"
      ],
      "metadata": {
        "id": "8xp_LptwMi-l"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "scaler = StandardScaler()\n",
        "train_scaled = scaler.fit_transform(X_for_ML)\n",
        "test_scaled = scaler.transform(X_test_ML)\n",
        "d_train = lgb.Dataset(train_scaled, label=train_labels_encoded)"
      ],
      "metadata": {
        "id": "hnjZ2OnWIsIk"
      },
      "execution_count": 255,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "lgbm_params = {'learning_rate':0.05, 'boosting_type':'gbdt', #Sometimes dart gives better accuracy. gbdt means gradient boosting.\n",
        "              'objective':'binary',\n",
        "              'metric':['auc', 'binary_logloss'],\n",
        "              'num_leaves':100, # optimizing this avoids overfitting. 50 to 100\n",
        "              'max_depth':10} #10 may be appropriate"
      ],
      "metadata": {
        "id": "PQspjVoaI_Vm"
      },
      "execution_count": 256,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "clf = lgb.train(lgbm_params, d_train, 50) #50 iterations. Increase iterations for small learning rates. eg if learning grate = 0.005, we may need 500 iterations.\n",
        "prediction=clf.predict(test_scaled)\n",
        "#convert the probabilities into binary values 0/1 for classification\n",
        "for i in range(0, test_scaled.shape[0]):\n",
        "    if prediction[i]>=.5: # setting threshold to 0.5\n",
        "       prediction[i]=1\n",
        "    else:  \n",
        "       prediction[i]=0"
      ],
      "metadata": {
        "id": "GrvqxgowJDZ-"
      },
      "execution_count": 257,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#Print accuracy\n",
        "print (\"Accuracy with LGBM = \", accuracy_score(test_labels_encoded, prediction))\n",
        "#Confusion matrix\n",
        "cm_lgbm = confusion_matrix(test_labels_encoded, prediction)\n",
        "sns.heatmap(cm_lgbm, annot=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 304
        },
        "id": "eEpCBvMnJKvK",
        "outputId": "f06ee6dc-fdac-4a20-cfdd-1f05ed8ef637"
      },
      "execution_count": 258,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy with LGBM =  0.5463917525773195\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f2a958aea50>"
            ]
          },
          "metadata": {},
          "execution_count": 258
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "prediction_prob=clf.predict(test_scaled)\n",
        "print(\"Area under curve (AUC): \", roc_auc_score(test_labels_encoded, prediction_prob))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xcOhHaJWJ1d8",
        "outputId": "647f265b-db34-4be1-857a-0a3ae9b12439"
      },
      "execution_count": 259,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Area under curve (AUC):  0.5317283163265306\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr, tpr, thresholds = roc_curve(test_labels_encoded, prediction_prob) # outputs false positive rate, true positive rate, and corresponding thresholds.\n",
        "plt.figure(1)\n",
        "plt.plot([0,1],[0,1], 'y--')\n",
        "plt.plot(fpr, tpr, linestyle = '--')\n",
        "plt.xlabel('False positive rate')\n",
        "plt.ylabel('True positive rate')\n",
        "plt.title('ROC curve')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "x2VvGptHKCJV",
        "outputId": "309db92c-da47-4ce5-a959-1cfde9e7a08f"
      },
      "execution_count": 260,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3xV9f348dc7N+NmD8IMCSGDALLUKOAEV93W1m1trSjWVVdrXV+l1G+rVuuo2opb61794mjVVq3+rCJLQNCQMJNAIGzIzs3798c5YMjiCrn3Jve+n49HHpzxOee8D4H7vp8z3h9RVYwxxkSuqFAHYIwxJrQsERhjTISzRGCMMRHOEoExxkQ4SwTGGBPhLBEYY0yEs0RgjDERzhKBCTsislJE6kRkh4hUicjTIpLUps0hIvKhiGwXka0i8paIjGzTJkVE7heR1e6+lrnzmcE9I2MCyxKBCVenqGoSMA7YH7hp5woRmQi8D/wfMAgYCiwAPhORPLdNLPBvYD/geCAFmAhsBA4OVNAiEh2ofRvTGUsEJqypahXwHk5C2Olu4FlVfUBVt6vqJlW9FfgCmOa2+SmQA5yuqktUtUVV16vq71T13Y6OJSL7icgHIrJJRNaJyM3u8qdF5I5W7SaJSEWr+ZUi8hsRWQjUuNOvtdn3AyLyoDudKiJPiMhaEakUkTtExLOPf1UmglkiMGFNRAYDJwBl7nwCcAjwagfNXwGOdaePAf6pqjv8PE4y8C/gnzi9jAKcHoW/zgVOAtKAl4AT3X3ifsifBbzgtn0aaHaPsT9wHHDx9ziWMbuxRGDC1d9FZDtQDqwHbneXZ+D8u1/bwTZrgZ3X//t00qYzJwNVqnqvqta7PY1Z32P7B1W1XFXrVHUVMA843V13FFCrql+ISH/gROAaVa1R1fXAfcA53+NYxuzGEoEJVz9U1WRgEjCc7z7gNwMtwMAOthkIbHCnN3bSpjPZwLK9itRR3mb+BZxeAsB5fNcbGALEAGtFZIuIbAEeBfrtw7FNhLNEYMKaqv4H51LKPe58DfA5cGYHzc/iu8s5/wJ+ICKJfh6qHMjrZF0NkNBqfkBHobaZfxWY5F7aOp3vEkE50ABkqmqa+5Oiqvv5Gacx7VgiMJHgfuBYERnrzt8I/ExEfikiySKS7t7MnQj81m3zHM6H7usiMlxEokSkj4jcLCIndnCMt4GBInKNiMS5+x3vrvsK55p/hogMAK7ZU8CqWg18DDwFrFDVb9zla3GeeLrXfbw1SkTyReTIvfh7MQawRGAigPuh+ixwmzv//4AfAD/CuQ+wCuem62GqWuq2acC5Yfwt8AGwDfgS5xJTu2v/qrod50bzKUAVUApMdlc/h/N46kqcD/GX/Qz9BTeGF9os/ykQCyzBudT1Gt/vMpYxuxEbmMYYYyKb9QiMMSbCWSIwxpgIZ4nAGGMinCUCY4yJcL2uwFVmZqbm5uaGOgxjjOlV5s6du0FV+3a0rtclgtzcXObMmRPqMIwxplcRkVWdrbNLQ8YYE+EsERhjTISzRGCMMRHOEoExxkQ4SwTGGBPhApYIRORJEVkvIl93sl5E5EERKRORhSJyQKBiMcYY07lA9giexhn0uzMnAIXuz1TgLwGMxRhjTCcC9h6Bqn4iIrldNDkNZwBxBb4QkTQRGejWWzfGmIjS2NzCQx+Wtls+Pq8PE4bG09hYTXx8bkCOHcoXyrLYfXi+CndZu0QgIlNxeg3k5OQEJThjjAmm5pYW/vxRWbvl9fUr8Wy4nOjoVA48cA4i3X8hp1e8WayqM4AZAMXFxTaAgjGmV1u/rZ6ZC9agCvPLN/PBknX898ajWfGHk3a1aWrawvLlv2bt2scRKaCg4L6AJAEIbSKoxBnwe6fB7jJjjAlrz89azQP//u4yUFpCDElx330cq/qYP/8QamtLyM6+gdzcaXg88QGLJ5SJYCZwpYi8BIwHttr9AWNMuFq7tY5PllbTP8XLIfl98MZ4uGDiEABiPVHERkfR1LSR6OgMRDwMHfq/xMVlk5JSHPDYApYIRORFYBKQKSIVwO1ADICq/hV4FzgRKANqgZ8HKhZjjAm1B/9dxotfrubwwkyemzKe8Xl9dq1TVaqq/kZZ2dXk5d3JoEGX0Lfv6UGLLZBPDZ27h/UKXBGo4xtjTE/S5Guhf0oc9509brfl9fXlLF36CzZtepeUlAmkph4a9Nh6xc1iY4wJB9FRUWQmxe2aX7fuRZYuvRRVHwUF95OVdSUinuDHFfQjGmOMASA6Op2UlPEMGzaD+PihIYtDnCs0vUdxcbHawDTGmN7m68qtbNheR17Cc6g2MmTILYBzf0BEAn58EZmrqh3eebYegTHGBMiOhmbK1u8AoL6ulKQdl7G8ai59+561KwEEIwnsiSUCY4wJkBteW8C7i6oAiPPU8eSJ5Ywc+Sp9+/64RySAnSwRGGNMN6ptbCYu2oMnSli9qZahfaI5Pfd2MtKP4OCDlxAT02fPOwkySwTGGNONLnxyNr//YT7J8gF9EgtJjovhkpNfID4+L9ShdcoSgTHG7KP6Jh/b6psAqNq6if959QMuHT2Nq46cR2ZqIfHxiSGOsGuWCIwxZh9NeWY2n5Vt3DWfnQjjxv2HtLRxXWzVc1giMMaYffTUhQfypzcuprZ+IxkZx3Ha+KtIS0sPdVh+s0RgjDF7qbFxAzExGcRGxzBl8ul4vTkkJ/e+UXdt8HpjjPmenCJxz/Lll8NYu/ZxrnhhHh8sP6BXJgGwHoExxnwv9fWrWPLNL7ji7dNYW/McMR5o8q1lQIo31KHtNUsExhjjp6qqv1FaehkNzdGkJV1MYmIix44cgABnHDg41OHtNUsExhjjp9jYvqSkHEpR0aMcM3lIqMPpNpYIjDGmEy0tTZSX34tqE7m5/0NGxg9ITz+uR5WH6A6WCIwxpgPbt8+npGQKO3bMJy7lZ1z46ke71tU3tVC1rZ47fzSacw7OCWGU3cMSgTHGtOLz1bNq1XRWr74bjRpAn+zXGdjvJMYuXbJbO0+UcPDQjBBF2b0sERhjTCt1dWWUl9/DgAE/Zb3cxqmPLeblqTt44Jz9Qx1awNh7BMaYiNfcvIOqqucASEoaxcEHlzB8+JNEe5JCHFlwWI/AGBPRNm16j5KSqTQ0lJOcXExi4gji44cy9dk5VGyuC3V4QWE9AmNMRGpq2sg33/yMhQuPx+NJYP/9PyUxccSu9c0tSmKch0lFfRnWPzmEkQae9QiMMRFH1ce8eYdSV1fG25Uz2OYbi5REAXPYXt/MWcXZPHzeAcTHekIdalBYIjDGRIzGxmpiYvog4iE//y7i4obwSqWP1ZtqdrVRhVfmlHP8qAEhjDS4LBEYY8KeUyTuaZYtu468vDsZNOhS5q4bz5I123jk/GGhDi/k7B6BMSas1dWtZOHCH1BSchGJiaNJS5sMwH/LNvDM5ytDGltPYT0CY0zYqqp6jqVLL0NEKCx8hEGDLkXEvv+2ZYnAGBO2YmP7k5Z2BMOG/RWvt/eXgggUSwTGmLDhFIm7G1Ufubm3kZFxHBkZx4U6rB7PEoExJixs3z6Pb7+9iJqaBfTrdx6qGnZVQgPFEoExplfz+epYufK3vDX3YzY0jKFv3+kkbhxFdkMzyd4YvirfwpyVm9pt95sThnPbKfuFIOKeJ6CJQESOBx4APMDjqnpnm/U5wDNAmtvmRlV9N5AxGWPCS339cioq/sSf57+OT3feCP6G40cNINkbw2dlG/jjeyXttvvxAYNJiLXvwgCiqoHZsYgHWAocC1QAs4FzVXVJqzYzgPmq+hcRGQm8q6q5Xe23uLhY58yZE5CYjTE9Q01DM+8trqLZpxw+LJOBqfGUb6rl82UbAfC11LF9+zzSUg9l8vB+JMdUMeZ3S/j5YblcMbkAgKTYaKKihPomH42+lnbH2Lk+UojIXFUt7mhdINPhwUCZqi53g3gJOA1oXdRbgRR3OhVYE8B4jDG9xMwFa7jpjUUAPP3zgxiYGs/iNdu44fWFrVolAwt57RcTKc4dwm9OaCHWI6R4Y3bblzfGgzcmMkpF7K1AJoIsoLzVfAUwvk2bacD7InIVkAgc09GORGQqMBUgJ8ceATOmN1uzpY4F5VvaLR+VlUp2RgJbahv5arWz/u2rDqOgn1MKemKu8MyPZ7Jhw5vExxeSl3c3ySkHkJkUC8D543OIi7Z3BPZGqC+QnQs8rar3ishE4DkRGaWqu/XjVHUGMAOcS0MhiNMY000+/HY9j326nFUba3dbvnPYx5Uba3l5Tjmx0VHk9U3EG+NB1ce3i4+A+uUcMPxmhgy5maiouN22t2/9ey+QiaASyG41P9hd1toU4HgAVf1cRLxAJrA+gHEZY7rJ/yvd0G7ZoDQveX2TaPa18MXy9k/rHFHYlwl5GTS37P6dbkCKF4Ci/sn885rDyUiIJZqNqPZ1i8Tdg9c7hKSkMYE5mQgWyEQwGygUkaE4CeAc4Lw2bVYDRwNPi8gIwAtUBzAmY0w3+skTs9otu/iwodx68kjqm1s6XH/10YVce2znhd7iYz0U9U+mqupJZi24nry8O8nK+gWZmad0a+zmOwFLBKraLCJXAu/hPBr6pKouFpHpwBxVnQlcDzwmItfi3Di+UAP1GJMxpluUVG3n0f8s48qjCnj1FxPbre+f7Hyz90ZHdbh+UFp8l/uvq1tOScklbNnyIampR5Ke3uGtQ9ONAnqPwH0n4N02y25rNb0EODSQMRhjulf19gbemF/JueNzOCg3o9N20Z6oLtd3pKrqGZYuvRwRD8OG/ZWBAy+xInFBEOqbxcYYs0ts7CDS04+isPAveL2DQx1OxLBEYIwJmZaWRlavvhPVFoYOnUZGxrFkZBwb6rAijiUCY8JYfZOP7fXNAIhAZpLzyOXWuiYam3d/2zZKoM/O9bVN7d7G9UQJGYmx3Rbbtm2zKSm5iJqar+nf/wIrEhdClgiMCVOqyhF3f8T67Q0ApCXE8NVtTknmG19fyD++rtqtfVZaPJ/deBQAV744j0/bPBo6rH8S7197JIPT4xmamUiMZ++u3ft8taxYcRsVFfcRGzuQUaNm2hNBIWaJwJgwpQrrtzdw1PB+TB7ej7hWH9znHJzDIQWZu7VPjP3uhawLD8nluP12H7w9Ld4p3ZCbmcjvThvF6KzUvYqrvn4FlZV/ZuDAS8jPv4vo6L3bj+k+lgiMCXNjB6dxwYQhuy07cljfLrc5ekT/LtcfVpjZ5fq2mpu3Ul39BgMH/pzExP0YP74Mrzd7zxuaoLBEYEyYEoGSO47HE+Lr7hs3vkNJyaU0Nq4lJWUiiYnDLQn0MPaArjFhSkSIi/YQvZfX8vdVY2M1S5acz6JFJxMTk84BB3xOYuLwkMRiumY9AmPCVEuLMv3tJUwq6sukon5BPbaqj/nzD6O+fgW5ub8lJ+dGoqK674kj070sERgTxp7+70rSE2KDlggaGqqIje3nFom7F683l6SkUUE5ttl7dmnIGLPPVFtYs+ZRvvxyGGvWPApAZubJlgR6Cb96BCISD+SoavuBP40xQbesegdXPD+PuiYfUa1uBt9z5hgOHJLBR9+u57dvLQ5KLLW1ZSxdeglbtnxMWtpRZGT8ICjHNd1nj4lARE4B7gFigaEiMg6YrqqnBjo4Y0zHaht8eGM8pMTH7KrjD5AU5zzrn5oQw5jBaeyfk87xowZ0tpt9tnbtU5SWXo5ILMOGPcbAgVPs7eBeyJ8ewTSc8Yc/BlDVr9wxBowxQbSoYiu3z/ya5hbl6qML+fsVnRfuPSAnnQNy0gMek9ebQ3r6Dxg27GHi4rICfjwTGP4kgiZV3domy9uYAcYE2ZxVm5i3eguHFWQSFx2aYRlbWhpYteoPQAtDh04nPf1o0tOPDkkspvv4kwgWi8h5gEdECoFfAv8NbFjGmNaqtzfw7OerAHjovP1JSwj+o5jbts3i22+nUFu7mP79f2ZF4sKIP08NXQXsBzQALwBbgasDGZQxxrFqYw2zlm8EoE9iLKeOHUSKNyaoMfh8NZSVXce8eRPx+bYyevTbjBjxtCWBMOJPj+AkVb0FuGXnAhE5E3g1YFEZE8E+WVrNC7NWA1CybjtrttRRcscJvHbZISGJp75+FZWVjzBo0C/Iy7uT6OiUkMRhAsefHsFNfi4zxnSDN+ZV8O9v17FiQw2xnihOGj0w6DE0NW1hzZrHAUhMHMn48WUMG/aIJYEw1WmPQEROAE4EskTkwVarUoDmQAdmTKS6/rgiLp9cwLD+ySE5/oYN/8fSpZfR2Lie1NTD3CJxNmxkOOvq0tAaYA5wKjC31fLtwLWBDMqYSJadkRCS4zY2rqe09JdUV79MYuIYRo2aaUXiIkSniUBVFwALROQFVW0KYkzGhK1X5pRTUrWdnIwEfnZILgCPfFzGxh2Nu9qUrt/Bj/bP4of7B++5fKdI3KHU169m6NA7yM6+gaio4N6UNqHjz83iXBH5AzAS2PUKo6rmBSwqY8LUtJmLafK1MDE/c1ci+MeiKlZsqNnVpsnXQlp8TFASQUPDGmJjByDioaDgAbzeXBITRwb8uKZn8ScRPAXcDtwHTAZ+jhWrM6ZTLS3Ks5+vZEtdEyneGC46zHkR/415FTQ2t3DRYUO5+cQRu9q/ddVhQY9xZ5G45ct/Q17enWRlXU6fPicGPQ7TM/iTCOJV9d8iIqq6CpgmInOB2wIcmzG90vINNUx7awkA2RnxrRJBJT5V8jITQxketbVLKSm5hK1bPyE9/RgyMk4IaTwm9PxJBA0iEgWUisiVQCWQFNiwjOm9WtSpwPLQefvv9ujnc1MOBgjpi1hr1z5BaemVREV5KSp6kgEDLrQXw4xfieBqIAGntMTvcC4P/SyQQRkTDgTZ7UO2J3zger25ZGScQGHhw8TFBf/9BNMzdZkIRMQDnK2qvwJ24NwfMMZ0IT0hlisnF1DQL/Qd55aWBlau/B0AeXl3WJE406EuE4Gq+kQk+HeyjOnF+ibH8asfFIU6DLZu/S8lJVOorf2WAQMusiJxplP+XBqaLyIzcWoL7XrGTVXfCFhUxvRizb4WttQ1kRQXjTcm+OWim5t3sGLFLVRW/pm4uGzGjPmnjRpmuuTPY6BeYCNwFHCK+3OyPzsXkeNFpEREykTkxk7anCUiS0RksYi84G/gxvRUyzfUUHzHv/j3N+tDcvyGhtWsWfMoWVlXcNBBX1sSMHu0xx6Bqu7VfQH3/sLDwLFABTBbRGaq6pJWbQpxCtgdqqqbRaTf3hzLmJ6gams9y6p3ULG5NujHbmraTHX1qwwaNJXExJFMmLCcuLhBQY/D9E5+DV6/lw4GylR1OYCIvAScBixp1eYS4GFV3QygqqH5CmVMN/ioZD03vbFo13ySN5D/vb5TXf0mpaWX09hYTVrakSQkFFkSMN9LIP+lZgHlreYrgPFt2gwDEJHPAA8wTVX/2XZHIjIVmAqQk5MTkGCN2VdHD+/HK5dOBMAbE8WoQakBPV5DQxVlZVdRXf0aSUnjGD36HRISQn+T2vQ+wfnK0vXxC4FJwGDgExEZrapbWjdS1RnADIDi4mIbL9n0OFtrm1ixoYYRg1KCMoKYqo+vvjqc+vpyhg79PdnZv7IicWav7fFmsYj0F5EnROQf7vxIEZnix74rgexW84PdZa1VADNVtUlVVwBLcRKDMb3KgootnD3jC0rXbQ/ocerrK1BtcYvEPUhx8VcMGXKTJQGzT/x5auhp4D1g50XHpcA1fmw3GygUkaEiEgucA8xs0+bvOL0BRCQT51LRcj/2bUxEUW2houLPfPnlcNas+QsAffqcYOMFmG7hTyLIVNVXgBYAVW0GfHvayG13JU4S+QZ4RVUXi8h0ETnVbfYesFFElgAfAb9W1Y17cR7GhK2amm+ZP/8Iysp+SWrqYfTp49fT28b4zZ97BDUi0gdQABGZAGz1Z+eq+i7wbptlt7WaVuA698cY08aaNY9TWnolHk8Cw4c/Q//+F9jbwabb+ZMIrse5pJPvPt3TFzgjoFEZE2KqSpPPeS4hSiDaE7XbstY8UYH7YI6Pzycz8xQKCx8iNrZ/wI5jIps/L5TNFZEjgSJAgBIbutKEuzve+YYn/t8KAC49Io+bThzB9oZmxkx7v13ba48ZxvkTcphxwYHk9923QnM+Xz2rVk0HIC/v96SnTyY9ffI+7dOYPdljIhCRhcBLwMuquizwIRkTeis31NAvOY6fHZLL/tlpAMR6ovh1B8XkDsrNIDMpjuP2G7BPx9y69TO+/XYKdXUlDBx4sRWJM0Hjz6WhU4CzgVdEpAV4GefG7+qARmZMiPVLieOKyQW75r0xnt3mu0tz83ZWrLiZysqH8XqHMGbMe2RkHNftxzGmM3t8akhVV6nq3ap6IHAeMAZYEfDIjAmhoZmJDOufHJRjNTRUsHbt42RlXUVx8SJLAibo/HqzWESG4PQKzsZ5dPSGQAZlTLBc8fw8/rtsA56o774THVrQhwfO2T+gx21q2sj69a+QlXUZiYkjGD9+uY0YZkLGn3sEs4AYnPEIztxZRM6YcDCsfzIrNtQwLidt17KiAPYEVJXq6tcpLb2C5uZNpKcf5RaJsyRgQkdUuy7dIyJFqloSpHj2qLi4WOfMmRPqMIz53hoa1lJaegUbNrxJUtKBDB/+BElJY0MdlokQIjJXVYs7Wtdpj0BEfqKqfwNOEpGT2q5X1T91Y4zGhMSW2kZiPFEkxgW2/qKqj/nzD6exsZK8vLsZPPhaoqJCXfPRGEdX/xIT3T876idbBVDTK6kq172ygOXVOwBYULGV88fn8L+njw7I8erry4mLy0LEw7BhD+P1DiUhYVhAjmXM3uo0Eajqo+7kv1T1s9brROTQgEZlTID4WpQ351cyNDORIX0SmFzUlxNHd//1eVUflZUPs3z5TeTn301W1hU2ZKTpsfzpm/4ZOMCPZcb0eFEi3HPmWEYOTGHkoJSAHKOm5htKSqawbdvnZGScQJ8+pwTkOMZ0l67uEUwEDgH6ikjronApOKOJGdPrREUJZxw4OGD7X7NmBqWlV+HxJDN8+HP073++vR1seryuegSxQJLbpvV9gm1Y0TnTi/halGkzF7NhRwOqUDQgmbMPymZQWny3Hys+vpDMzNMpLHyQ2Nh+3b5/YwKhq3sE/wH+IyJPq+qqIMZkTLdau7WO575YRf+UOFK8MazcWMP4vIxuSQQ+Xx0rV04DhPz8O61InOmVuro0dL+qXgM8JCLtnhJS1VM72MyYHuv644o4qzh7zw39tGXLJ5SUXExdXSmDBv3CisSZXqurS0PPuX/eE4xAjAmUvslxvHn5IeRkJHTL/pqbt7F8+Y2sWfMXvN48xo79N+npR3XLvo0Jha4uDc11//zPzmUikg5kq+rCIMRmTLeIi/awf056t+2voWENVVVPM3jwdQwdOh2PJ3HPGxnTg+2x+qiIfCwiKSKSAcwDHhMRe6vY9Bo1Dc28PHv1rpfI9kZj4wYqKx8BIDFxOBMmrKCg4F5LAiYs+DN4faqqbgN+BDyrquOBYwIbljHdZ3NtI795fRFzVm3+3tuqKuvXv8zs2SMpK7uG2tqlADZspAkr/iSCaBEZCJwFvB3geIzpMRoa1vD11z9kyZJz8HqHcOCBc608hAlL/rxZPB14D/hMVWeLSB5QGtiwjAktp0jcETQ2VpKffw9ZWVdbkTgTtvwZvP5VnLEIds4vB34cyKCMCZX6+lXExQ12i8Q9gtebR0JC9w9PaUxP4s/ANINxagvtLDT3KXC1qlYEMjBjvq+GZh/vL15HfZNv17Kx2WkkxO65Ioqqj4qKB1ix4la3TPSVNmSkiRj+9HWfAl4AznTnf+IuOzZQQRmzNz5ZuoGrXpy/27JbTxrBTyfm8sG1R9Avxdvhdjt2fE1JyRS2b/+SPn1OJjPzh8EI15gew59E0FdVn2o1/7SIXBOogIzZW4Lz8tgDZ48j2315LDUhhtjoKAo7GX6ysvKvlJX9kujoVEaMeIF+/c6xt4NNxPEnEWwUkZ8AL7rz5wIbAxeSMXvnmJH9OWakf4917iwHkZg4gr59z6Sg4H5iY/sGOEJjeiZ/EsFFOPcI7nPnPwN+HrCIjPkemnwtzF+9hWZfCwDDB6aQkRjbaXufr5YVK25DxEN+/l2kpR1JWtqRwQrXmB7Jn6eGVgFWYM70SK/PreDGNxbtmp9U1JcHzt6f1ISYdm03b/6YkpKLqa9fxqBBl1uROGNc/jw1lAc8AEzAGav4c+Ba9zFSY0KqptF5QujJC4tJjI0mLSG2XRJobt7KsmU3sHbtDLzefMaO/dBKRRvTij+Xhl4AHgZOd+fPwblfMD5QQRnjr4GpXibkZTB+aB8S4zr+59zQsJZ16/5GdvavyM39LR5P91QhNSZc+FNiIkFVn1PVZvfnb0DHz+G1ISLHi0iJiJSJyI1dtPuxiKiIFPsbuDEAJ44eyEtTJ7ZLAo2N1VRU/BnYWSRuJfn5f7QkYEwH/OkR/MP9EH8J59LQ2cC7bjVSVHVTRxuJiAenJ3EsUAHMFpGZqrqkTbtk4Gpg1l6fhTEup0jci5SW/hKfbxsZGT8gIWGYPRFkTBf86RGcBVwKfAR8DFyGc3loLjCni+0OBspUdbmqNuIkktM6aPc74C6g3v+wjXG8OqecSX/8iO31TdTXl7No0Sl88835xMcXUFw834rEGeMHf54aGrqX+84CylvNV9DmvoKIHIAz0M07IvLrznYkIlOBqQA5OTl7GY4JR9vqm1m5sRafr5nFiybR2FhFfv59DB58FU6n1BizJyErpygiUcCfgAv31FZVZwAzAIqLi9uNn2wiU7OvhYYG591GiYpm2LBHiY/PIz4+L8SRGdO7BDIRVAKtRwof7C7bKRkYBXzsPss9AJgpIqeqaleXnIyhpaWZqU+9wodlqQB4ooTUDBsvyZi9EchEMBsoFJGhOAngHOC8nStVdSuQuXNeRD4GfmVJwOzJtu0LKCm5hP2Soxk09jhG559BUiePjhpj9syfF8oEOB/IU9XpIpIDDFDVL7vaTlWbReRKnEFtPMCTqrpYRKYDc1R1ZjfEbyLAXf/8lqc+WwGAajMNzVFER93Mf69ppAOEYVMAABaDSURBVG/fM+3tYGP2kT9fox4BWoCjcEYr2w68Dhy0pw1V9V3g3TbLbuuk7SQ/YjERZEH5FkSgeEg6vhZFgPqGcmp2LOTgopPo129EqEM0Jiz4kwjGq+oBIjIfQFU3i0jnVb2M2UsvzFrNff9aumt+S20jBw5J5Y6jXiO3KJr8/D8CIwAbMMaY7uRPImhyXw5TABHpi9NDMKZb3PNeCQfmpjOkTwLHjPiujHRjQyUFCX+gouIdsrKusiJxxgSIP4ngQeBNoJ+I/C9wBnBrQKMyYa3Z18KFT82mcksdIrC8uoZfHJnPjScM59CCTJqatrBs2a+oqnqC+PhCioo+IS3t8FCHbUzY8ueFsudFZC5wNM4gUD9U1W8CHpkJWzsampm9chONvhZOGj2QUYNSOWn0wF3rm5rWsX79S2Rn/4bc3NvxeOJDGK0x4U9Uu34/y31KqB1VXR2QiPaguLhY58yxJ0zDTWOj8+E/ePDV7vwGYmMz97CVMcZfIjJXVTss7OnPpaF3cO4PCE7V0aFACbBft0VoIpaqsm7d85SVXY3Pt4OMjBNJSCi0JGBMEPlzaWh063m3PtDlAYvIhL3t9U1c98oCfjzOyyBuYNOmf5CSMpGioidISCgMdXjGRJzv/Tqmqs4TERuUxuy1Zp/ywZJ1ZHle5ajsTygoeJCsrMutSJwxIeLPm8XXtZqNAg4A1gQsIhO2ttQ2cuNrX9Cszs3ffv3P5aCDphEfnxvawIyJcP70CJJbTTfj3DN4PTDhmHDV0tLMqtUPsnBVErGx/Rk+oA8Th40mPj491KEZE/G6TATui2TJqvqrIMVjwtD27V9RUjKFHTvmMeP00yksfJi4uIF73tAYExSdjlAmItGq6gMODWI8JsxUVDzEvHkHsWJjM7fPmsmm2EctCRjTw3TVI/gS537AVyIyE3gVqNm5UlXfCHBsphdTVWobfTw7v4gNm+9CYw9j1eZqahubQx2aMaYNf+4ReIGNONVHd75PoIAlAtNOc/MOVqy4BZEYNPl2Fq5N4ovlRSTEbiIjMZYhfRJDHaIxpo2uEkE/94mhr/kuAexkw0WadjZtep+Skqk0NKwmK+sqCvol8dLUiaEOyxizB10lAg+QxO4JYCdLBGaXpqbNLFt2HVVVTxMfX8S4cZ+QlnZYqMMyxvipq0SwVlWnBy0S02s1Na2nuvo1cnJuIqP/zby9aCNNvhWUrt/BC7NW8+blh7B/jj0makxP1VUisMLvplMNDVWsX/8i2dnXkpBQxIQJK4mJ6cNrcyu49e9f72oX64kiMykuhJEaY/akq0RwdNCiML2GUyTuWcrKrsXnq6VPn5NJSCgkJqYPAKOyUrj1pBGcOHogCbEe4qI9xMda6QhjerJOE4GqbgpmIKbnq6tbydKll7J58/ukpBxKUdHj7YrEDR+QwvABKSGK0BizN7530TkTmVpamvl41ml8W53BgP5PcsS4nxIV5WHxmq1UbK7b1W5bXRN5fZMYMziVGE+n7ysaY3oQSwSmS7W1ZcTHDyUqKprXV9zFv0p8AJxzuPMh//ys1bwwa/cxijxRwrz/OZbUeEsExvQGlghMh1pamigv/yMrV/6W/Pw/ogkXs6E2mcJ+zdx/zrhd7a6cXMD543cfxC49IZbU+Jhgh2yM2UuWCEw727fPc4vEfUXfvmfSr9/Z/PXTSr4q38LEvD7sNyh1V9tBafEMSrMxhY3pzSwRmN0sLn2IOd88QExMOjnZb/LC0nyyqrdz+v5ZHJSbQWH/pFCHaIzpZpYIDOA8FioizF47gmn/vb/VmtWcOnYQl09KIDsjIWTxGWMCxxJBmKpv8rF+W0O75X2SYkmMi6au0Uf19gaafTtYseperpp5CNceO5oTxxzCoD5bd9vG3go2JrxZIghTS9Zu40eP/Lfd8ofPO4CTxgxkzqpNXPDEl+7SI9w/lay0eLLsmr8xEcUSQRjZWttEizr1ALPS4rn3zLHt2owZnEpT00Zk261cMrqM2NhBDBw4heSkQiYP7xfskI0xPYAlgjDx+KfLueOdbwDwxkRx5eQCLptUgCeqfcmo2toKtPZFzjvkGoYMuYWoKKsFZEwkC2giEJHjgQdwSlo/rqp3tll/HXAx0AxUAxep6qpAxhSu1m6tJzY6iptPGA5AVnrCbkmgoWEt69Y9T3b29SQkDGPChFXExNi1f2NMABOBO/D9w8CxQAUwW0RmquqSVs3mA8WqWisilwF3A2cHKqZwFiUQH+PhwkOH7rZcVamqeoqysutQbSAz8zS3SJwlAWOMI5A1AA4GylR1uao2Ai8Bp7VuoKofqWqtO/sFMDiA8YS1W04ayYLbj9ttWV3dChYuPI6SkikkJY2luHhBuyJxxhgTyEtDWUB5q/kKYHwX7acA/+hohYhMBaYC5OTkdNQkom2qaaR8Uy1js9N2LWtpaWbBgqNoatpIYeFfGDRoKiJW+8cY016PuFksIj8BioEjO1qvqjOAGQDFxcU2TGYr/1i0lv/5v6+Ji/bw8a8n0dSwjPj4PKKioikqeor4+Hy83uxQh2mM6cEC+RWxEmj9CTTYXbYbETkGuAU4VVXbvwFlOvX4p8u57Pl5DEj1MuOCcVSW/57Zs0dRWfkQAOnpkywJGGP2KJA9gtlAoYgMxUkA5wDntW4gIvsDjwLHq+r6AMYSluaXbyErLZ5nfuJledlkqmsW0q/fOfTrd26oQzPG9CIB6xGoajNwJfAe8A3wiqouFpHpInKq2+yPQBLwqoh8JSIzAxVPuHjow1I+/HYdAMlx0XjYyqIFE2hq2sCoUf/HyJEvEhtrL4YZY/wX0HsEqvou8G6bZbe1mj4mkMcPRx+VVPNJ6QYmF/XjwkNzGZa5gYEDppCXdzcxMWl73oExxrTRI24Wm/ZuemMRE/IyOG1cFhWba7nqxfkALK3azrA+VSxb9grDC+5j+ICjgKNCG6wxplez5wl7qBe/XM1rcysAiBIhKS6auKgt5KcupDjzb4hEo2oPUBlj9p31CHqoGI8wOssZCSwzoYabJv6Z9eufJyFhP4YPf4KUlK5eyTDGGP9ZIugFmps3s3HjWwwZcjtDhtxMVFRsqEMyxoQRSwQ92LZtX6BaREJCoVskzm4GG2O6n90j6GFUlTVrHqNf/Gp8te9SV7cMwJKAMSZgrEcQQo3NLdz7fgnb6psAaG7expYtnzLQ+ykPn1zHsGGPkZBQEOIojTHhzhJBCG2saWDFhhr+9c06MpPiaGxcB2RzeP6FjB17hhWJM8YEhSWCEBqYGs/9ZyTj9Y4jKiqaLVv+g9ebj9dr1biNMcFjXzlDpKWlkRUrfssXs8ZSUfkwAGlpR1oSMMYEnSWCENi27Uvmzj2QTxY9w0Xvvc6iTSeEOiRjTASzRBBk5eX3M2/eRJqaNlNQ8CAA0dEpIY7KGBPJLBEEyc5yECkpBzNw4CUcfPBi0tIOD3FUxhhjN4sDrrl5K8uW3UBUVDyFhfeTmnoIqamHuGu3hjQ2Y4wB6xEE1IYNb/HllyNZu/ZxoqLirEicMaZHsh5BADQ2VlNWdjXr179IYuJoRo36OykpB7Vrl5kUx6VH5jE0MzEEURpjjMMSQTfaUtvI7JWbaWhYy7JlVWRm3s8PR08l0RvP8uodLKuuabfN9ccWERttHTNjTOhYIugm9fXlTH/zQ95YlOkuuR6AI8cqiV54d9Fa7nl/abvtvrrtWGKjrZqoMSZ0LBHsI9UW1qyZwfLlN7Bx88VkJBzHs1Mm7lrfNzkOgLOKs5lU1H4s4aQ4+xUYY0LLPoX2QW1tKSUll7B1639ISzuaRy66Aq83D0+UtGvbL8VLvxRvCKI0xpiuWSLYSy0tzSxYcCzNzVsoKnqCAQN+jkj7BGCMMT2dJYLvqabmG+LjC4mKimbEiOeIj88nLm4QAG/Or6B03Q5uOH54iKM0xhj/2eMqfmppaWDFituZM2cMlZUPAZCWdviuJADwxbJNvDGvMlQhGmPMXrEegR+2bv2CkpIp1NYuoX//Cxgw4IJQh2SMMd3GEsEelJffy7JlvyYmNpucgnfIyDiGmiagqREAb4wHb4wHX4vS0OwLbbDGGLMXLBF0QrUFkShSUiYyaNAvSO47nYl3zQI+2K3dTScM59Ij81m9qZa/f7WGwenxoQnYGGP2UkQmAl9L+5o/AkRFCU1NW1haej0eTwKFhQ+SlDyRpOSJ1DX5uP2Uke22Kx6SAUBGYiy3nzKSkQOtpLQxpneJyESw//T32VbfvNuyMw8czA2TVlBaejk/eeuvtKgHeHfX+gsPyWXaqft1us/U+Bh+fujQQIVsjDEBEzGJ4MNv1/H2grX84cejuWJyAQ3NLbvW+Xw1pLQ8y+LFfyIpaRyXH5FKbOyg3bYfm50W7JCNMSYoIiYRlK7bwRvzK7nj9FFcemT+butqa8uYN+9JsrP/l+zsX1McFROiKI0xJvgiJhG0VV+/mnXrniMn52YSEgqYMGE10dHJoQ7LGGOCLqAvlInI8SJSIiJlInJjB+vjRORld/0sEckNZDzgPA1UWfkIs2fvx6pVv6eubhmAJQFjTMQKWCIQEQ/wMHACMBI4V0TaPnYzBdisqgXAfcBdgYpnp4ULT6C09ApSUiZy0EGLSUgoCPQhjTGmRwtkj+BgoExVl6tqI/AScFqbNqcBz7jTrwFHS4Aqt8VFQ3LsdmpqFlNU9BRjxrxHfHxuIA5ljDG9SiDvEWQB5a3mK4DxnbVR1WYR2Qr0ATa0biQiU4GpADk5OXsVzIWH5vPD/dYSHz+fuLiBe7UPY4wJR72i6JyqzlDVYlUt7tu3717vJy3tMEsCxhjTRiATQSWQ3Wp+sLuswzYiEg2kAhsDGJMxxpg2ApkIZgOFIjJURGKBc4CZbdrMBH7mTp8BfKiq7es/GGOMCZiA3SNwr/lfCbwHeIAnVXWxiEwH5qjqTOAJ4DkRKQM24SQLY4wxQRTQF8pU9V1aF+xxlt3WaroeODOQMRhjjOlar7hZbIwxJnAsERhjTISzRGCMMRHOEoExxkQ46W1Pa4pINbBqLzfPpM1byxHAzjky2DlHhn055yGq2uEbub0uEewLEZmjqsWhjiOY7Jwjg51zZAjUOdulIWOMiXCWCIwxJsJFWiKYEeoAQsDOOTLYOUeGgJxzRN0jMMYY016k9QiMMca0YYnAGGMiXFgmAhE5XkRKRKRMRG7sYH2ciLzsrp8lIrnBj7J7+XHO14nIEhFZKCL/FpEhoYizO+3pnFu1+7GIqIj0+kcN/TlnETnL/V0vFpEXgh1jd/Pj33aOiHwkIvPdf98nhiLO7iIiT4rIehH5upP1IiIPun8fC0XkgH0+qKqG1Q9OyetlQB4QCywARrZpcznwV3f6HODlUMcdhHOeDCS405dFwjm77ZKBT4AvgOJQxx2E33MhMB9Id+f7hTruIJzzDOAyd3oksDLUce/jOR8BHAB83cn6E4F/AAJMAGbt6zHDsUdwMFCmqstVtRF4CTitTZvTgGfc6deAo0VEghhjd9vjOavqR6pa685+gTNiXG/mz+8Z4HfAXUB9MIMLEH/O+RLgYVXdDKCq64McY3fz55wVSHGnU4E1QYyv26nqJzjjs3TmNOBZdXwBpInIPo3BG46JIAsobzVf4S7rsI2qNgNbgT5BiS4w/Dnn1qbgfKPozfZ4zm6XOVtV3wlmYAHkz+95GDBMRD4TkS9E5PigRRcY/pzzNOAnIlKBM/7JVcEJLWS+7//3PQrowDSm5xGRnwDFwJGhjiWQRCQK+BNwYYhDCbZonMtDk3B6fZ+IyGhV3RLSqALrXOBpVb1XRCbijHo4SlVbQh1YbxGOPYJKILvV/GB3WYdtRCQapzu5MSjRBYY/54yIHAPcApyqqg1Bii1Q9nTOycAo4GMRWYlzLXVmL79h7M/vuQKYqapNqroCWIqTGHorf855CvAKgKp+DnhxirOFK7/+v38f4ZgIZgOFIjJURGJxbgbPbNNmJvAzd/oM4EN178L0Uns8ZxHZH3gUJwn09uvGsIdzVtWtqpqpqrmqmotzX+RUVZ0TmnC7hT//tv+O0xtARDJxLhUtD2aQ3cyfc14NHA0gIiNwEkF1UKMMrpnAT92nhyYAW1V17b7sMOwuDalqs4hcCbyH88TBk6q6WESmA3NUdSbwBE73sQznpsw5oYt43/l5zn8EkoBX3fviq1X11JAFvY/8POew4uc5vwccJyJLAB/wa1Xttb1dP8/5euAxEbkW58bxhb35i52IvIiTzDPd+x63AzEAqvpXnPsgJwJlQC3w830+Zi/++zLGGNMNwvHSkDHGmO/BEoExxkQ4SwTGGBPhLBEYY0yEs0RgjDERzhKB6bFExCciX7X6ye2i7Y7gRdY5ERkkIq+50+NaV8IUkVO7qpIagFhyReS8YB3P9F72+KjpsURkh6omdXfbYBGRC3Eqnl4ZwGNEu/WyOlo3CfiVqp4cqOOb8GA9AtNriEiSO5bCPBFZJCLtqo2KyEAR+cTtQXwtIoe7y48Tkc/dbV8VkXZJQ0Q+FpEHWm17sLs8Q0T+7tZ+/0JExrjLj2zVW5kvIsnut/Cv3bdgpwNnu+vPFpELReQhEUkVkVVuPSREJFFEykUkRkTyReSfIjJXRD4VkeEdxDlNRJ4Tkc9wXozMddvOc38OcZveCRzuHv9aEfGIyB9FZLZ7Lpd206/G9Hahrr1tP/bT2Q/Om7FfuT9v4rwJn+Kuy8R5s3Jnr3aH++f1wC3utAen5lAmzpgEie7y3wC3dXC8j4HH3OkjcOvBA38GbnenjwK+cqffAg51p5Pc+HJbbXch8FCr/e+aB/4PmOxOnw087k7/Gyh0p8fjlD9pG+c0YC4Q784nAF53uhDnjVtw3k59u9V2U4Fb3ek4YA4wNNS/Z/sJ/U/YlZgwYaVOVcftnBGRGOD3InIE0IJTerc/UNVqm9nAk27bv6vqVyJyJM6AJZ+55TVigc87OeaL4NSEF5EUEUkDDgN+7C7/UET6iEgK8BnwJxF5HnhDVSvE/2EtXsZJAB/hlDh5xO2lHMJ3ZUDA+cDuyExVrXOnY4CHRGQcTvIc1sk2xwFjROQMdz4VJ3Gs8DdoE54sEZje5HygL3CgqjaJU1XU27qB+wF+BHAS8LSI/AnYDHygquf6cYy2N806vYmmqneKyDs4dV8+E5Ef4P8AODNxkloGcCDwIZAIbGmd/LpQ02r6WmAdMBbncm9nMQhwlaq+52eMJkLYPQLTm6QC690kMBloN+6yOGMxr1PVx4DHcYb8+wI4VEQK3DaJItLZt+az3TaH4VR13Ap8ipOEdt6A3aCq20QkX1UXqepdOD2Rttfzt+NcmmpHVXe42zyAc/nGp6rbgBUicqZ7LBGRsX7+vaxVp/7+BTiXxDo6/nvAZW5vCREZJiKJfuzfhDnrEZje5HngLRFZhHN9+9sO2kwCfi0iTcAO4KeqWu0+wfOiiOy81HIrTq3+tupFZD7O5ZaL3GXTcC43LcSp9rizhPk1bkJqARbjjPrWesjAj4AbReQr4A8dHOtl4FU35p3OB/4iIre6MbyEM05vVx4BXheRnwL/5LvewkLAJyILgKdxkk4uME+ca0/VwA/3sG8TAezxUWNcIvIxzuOWvXnMAmO+N7s0ZIwxEc56BMYYE+GsR2CMMRHOEoExxkQ4SwTGGBPhLBEYY0yEs0RgjDER7v8DE5DxfSR3Yv8AAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}